CN112654700B - Gene coding system for constructing and detecting bioactive agents - Google Patents

Abstract

The present invention relates to the field of genetic engineering. In particular, the invention relates to the construction of operons for the production of biologically active agents. For example, operons can be constructed to generate reagents that control the function of biochemical pathway proteins (eg, protein phosphatases, kinases, and/or proteases). Such reagents may include inhibitors and modulators useful in the study or control of phosphatase function associated with abnormalities in phosphatase pathways or expression levels. Fusion proteins, such as light-activated protein phosphatases, can be genetically encoded and expressed as light-switchable phosphatases. Provides a system for controlling phosphatase function in living cells or identifying small molecule inhibitor/activator/modulator molecules of protein phosphatases associated with cell signaling.

Description

Genetically encoded systems for constructing and detecting bioactive agents

Government Support Statement

This invention was made with government support under Award Nos. 1750244 and 1804897 awarded by the National Science Foundation. The government has certain rights in this invention.

Technical Field

The present invention relates to the field of genetic engineering. In particular, the present invention relates to the construction of operons for producing bioactive agents. For example, operons can be constructed to produce reagents for controlling the functions of biochemical pathway proteins (e.g., protein phosphatases, kinases and/or proteases). Such reagents may include inhibitors and regulators that can be used to study or control the phosphatase functions associated with abnormalities in phosphatase pathways or expression levels. Fusion proteins, such as light-activated protein phosphatases, can be genetically encoded and expressed as light-controlled switch phosphatases. Systems for controlling the phosphatase functions in living cells or identifying small molecule inhibitors/activators/regulator molecules of protein phosphatases associated with cell signaling are provided.

Background Art

Protein phosphorylation is implicated in cellular signaling because it controls, in part, the location and timing of cell differentiation, migration, proliferation, and ^death1-4 ; its dysregulation is associated with disorders such as cancer, diabetes, obesity, and Alzheimer's ^disease5-9 . Optical tools that exert spatiotemporal control over the activity of phospho-regulated enzymes in living cells could elucidate the mechanisms by which cells transmit, filter, and integrate chemical ^signals10,11 , reveal connections between seemingly disparate physiological processes (e.g., ^memory12 and ^metabolism13 ), and facilitate the identification of new targets for phospho-regulated therapeutics, a class of ^drugs14 . Therefore, there is a need to develop tools to control, reduce, or enhance the activity of phospho-regulated enzymes in living cells.

Summary of the invention

The present invention relates to the field of genetic engineering. In particular, the present invention relates to the construction of operons for producing bioactive agents. For example, operons can be constructed to produce reagents for controlling the functions of biochemical pathway proteins (e.g., protein phosphatases, kinases and/or proteases). Such reagents may include inhibitors and regulators that can be used to study or control the phosphatase functions associated with abnormalities in phosphatase pathways or expression levels. Fusion proteins, such as light-activated protein phosphatases, can be genetically encoded and expressed as light-controlled switch phosphatases. Systems for controlling the phosphatase functions in living cells or identifying small molecule inhibitors/activators/regulator molecules of protein phosphatases associated with cell signaling are provided.

In one embodiment, the present invention contemplates a gene operon comprising: a) providing; i) a first gene encoding a first fusion protein, the first fusion protein comprising a substrate recognition domain and a DNA binding domain or an anchoring unit of an RNA polymerase; ii) a second gene encoding a second fusion protein, the second fusion protein comprising an enzyme substrate domain and an anchoring unit or a DNA binding domain of an RNA polymerase; iii) a first DNA sequence comprising a binding site for the DNA binding domain; iv) a second DNA sequence comprising a binding site for the anchoring unit and for the RNA polymerase close to the first; v) a third gene encoding a first enzyme, wherein the first enzyme is capable of modifying the substrate domain, thereby changing the affinity of the substrate recognition domain; vi) a fourth gene encoding a second enzyme, wherein the second enzyme is capable of not modifying the substrate domain; vii) a reporter gene encoding at least one gene capable of having a detectable output when the RNA polymerase and the anchoring unit bind to the second DNA sequence binding site after the two fusion proteins are associated. In one embodiment, the substrate domain is a peptide substrate of a protein kinase. In one embodiment, the substrate domain is a peptide substrate of a protein tyrosine kinase. In one embodiment, the substrate domain is a peptide substrate of a Src kinase (protein tyrosine kinase). In one embodiment, the substrate recognition domain can bind to the substrate domain in its phosphorylated state. In one embodiment, the substrate recognition domain can bind to the substrate domain in its unphosphorylated state. In one embodiment, the DNA binding domain is a 434cI repressor and the DNA binding site is a binding sequence of a 434cI repressor. In one embodiment, the anchoring unit is an ω subunit of an RNA polymerase and the second DNA binding site is a binding site of an RNA polymerase. In one embodiment, the substrate domain is a peptide substrate of a protein kinase. In one embodiment, the operon further comprises a protein system. In one embodiment, the first enzyme is a protein phosphatase. In one embodiment, the first enzyme is a protein tyrosine phosphatase. In one embodiment, the first enzyme is a protein tyrosine phosphatase 1B. In one embodiment, the second enzyme is a protein kinase. In one embodiment, the second enzyme is a protein tyrosine kinase. In one embodiment, the second enzyme is a Src kinase. In one embodiment, the reporter protein that produces a detectable output is a LuxAB bioreporter (e.g., the output is luminescence). In one embodiment, the reporter protein that produces a detectable output is a fluorescent protein. In one embodiment, the reporter protein that produces a detectable output is mClover. In one embodiment, the reporter protein that produces a detectable output confers antibiotic resistance. In one embodiment, the antibiotic resistance is to spectinomycin. In one embodiment, the operon further includes a gene encoding a bait protein fusion, which includes: (i) a second enzyme substrate domain different from the first enzyme substrate domain and (ii) a protein that does not specifically bind to DNA and/or RNA polymerase, and a fifth gene encoding a third enzyme, wherein the third enzyme is active on the bait substrate domain. In one embodiment, both the first enzyme substrate domain (of the base system) and the second enzyme substrate domain (of the bait) are substrates of protein kinases. In one embodiment, both the first enzyme substrate domain (of the base system) and the second enzyme substrate domain (of the bait) are substrates of protein tyrosine kinases. In one embodiment, both the first enzyme substrate domain (of the base system) and the second enzyme substrate domain (of the bait) are substrates of Src kinase. In one embodiment, both the first enzyme substrate domain (of the base system) and the second substrate domain (of the bait) are substrates of protein phosphatases. In one embodiment, both the first enzyme substrate domain (of the base system) and the second substrate domain (of the bait) are substrates of protein tyrosine phosphatases. In one embodiment, both the first enzyme substrate domain (of the base system) and the second substrate domain (of the bait) are substrates of protein tyrosine phosphatases 1B. In one embodiment, the first enzyme is a light-regulated enzyme. In one embodiment, the first enzyme is a protein-LOV2 chimera. In one embodiment, the first enzyme is a PTP1B-LOV2 chimera. In one embodiment, the protein that produces a detectable output includes a protein that generates a toxic product in the presence of a non-essential substrate. In one embodiment, the additional protein is SacB, which converts sucrose into a non-structural polysaccharide that is toxic in Escherichia coli. In one embodiment, the operon further comprises an expression vector and a bacterial cell.

In one embodiment, the present invention contemplates a system for detecting an inhibitor of an enzyme, comprising: a) providing; i) an operon comprising a gene encoding an enzyme; ii) a bacterial cell; iii) a small molecule test compound; and b) contacting the bacteria with the operon such that the contacted bacteria are capable of producing a detectable output; c) growing the contacted bacteria in the presence of the test compound under conditions that allow for the detectable output; and d) evaluating the effect of the test compound on the detectable output. In one embodiment, the enzyme is a protein phosphatase. In one embodiment, the enzyme is a protein tyrosine phosphatase. In one embodiment, the enzyme is protein tyrosine phosphatase 1B.

In one embodiment, the present invention contemplates a method for evolving an inhibitor of an enzyme, comprising: a) providing: i) an operon comprising genes encoding an enzyme; ii) a library of bacterial cells, wherein each of the bacterial cells has at least one mutated metabolic pathway; b) growing the library of bacterial cells; and c) screening the library of bacterial cells for detectable output. In one embodiment, the operon further comprises an expression vector.

In one embodiment, the present invention contemplates a method for detecting a selective inhibitor of a first enzyme relative to a second enzyme, comprising: a) providing; i) a system comprising a bacterial cell library as described above; and ii) a small molecule test compound; b) growing the bacterial cell library in the presence of the test compound; and c) evaluating the effect of the test compound on a detectable output. In one embodiment, the system further provides an operon comprising a gene encoding a bait fusion protein, the bait fusion protein comprising; (i) a second enzyme substrate domain different from the first enzyme substrate domain and (ii) a protein that does not specifically bind to a DNA and/or RNA polymerase. In one embodiment, the operon further comprises an expression vector.

In one embodiment, the present invention contemplates a method for evolving a selective inhibitor of a first enzyme relative to a second enzyme, comprising: a) providing; a system comprising a bacterial cell library having a metabolic pathway with mutations as described herein; b) growing the bacterial cell library; and b) screening the bacterial cell library for detectable output. In one embodiment, the method further provides an operon comprising a gene encoding a bait fusion protein, the bait fusion protein comprising: (i) a second enzyme substrate domain different from the first enzyme substrate domain and (ii) a protein that does not specifically bind to a DNA and/or RNA polymerase. In one embodiment, the operon further comprises an expression vector.

In one embodiment, the present invention contemplates a method for evolving a light-switchable enzyme, comprising: a) providing; i) a system as described herein comprising a bacterial cell library having a mutated light-switchable enzyme; b) growing the bacterial cell library under at least two different light conditions; and c) comparing the difference in detectable output per cell between each of the two different light conditions. In one embodiment, the operon further comprises an expression vector.

In one embodiment, the present invention contemplates a method for evolving a light-operated switch enzyme, comprising: a) providing; i) a system as described herein comprising a bacterial cell library having a mutated light-operated switch enzyme; b) growing the bacterial cell library under a first light source in which activity is desired; c) subsequently growing the bacterial cell library from step b) in the presence of: (i) a non-essential substrate and (ii) a second light source in which activity is not desired; d) subsequently screening the survivors of step c) for mutated bacterial cells; and e) examining the activity of the mutated bacterial cells under the first light source and the second light source. In one embodiment, the method further comprises an operon comprising a gene encoding a bait fusion protein, the bait fusion protein comprising: (i) a second enzyme substrate domain different from the first enzyme substrate domain; and (ii) a protein that does not specifically bind to a DNA and/or RNA polymerase. In one embodiment, the operon further comprises an expression vector.

In one embodiment, the present invention contemplates a method for evolving selective mutants of an enzyme, comprising: a) providing; a system as described above comprising a bacterial cell library of an enzyme having mutations; b) growing the bacterial cell library and c) comparing detectable output between cells to identify the mutated enzyme. In one embodiment, the method further comprises an operon comprising a gene encoding a bait fusion protein, the bait fusion protein comprising: (i) a second enzyme substrate domain different from the first enzyme substrate domain; and (ii) a protein that does not specifically bind to a DNA and/or RNA polymerase. In one embodiment, the operon further comprises an expression vector.

In one embodiment, the present invention contemplates a method for evolving a substrate domain that is selective for an enzyme, comprising: a) providing; a method as described above comprising a bacterial cell library comprising a substrate domain fused to a DNA binding domain; b) growing the bacterial cell library in the presence of an inducer of the first enzyme and a non-essential substrate; c) subsequently growing the bacterial cell library from step b) in the presence of an inducer of the second enzyme; and d) subsequently screening the survivor bacterial cells to identify substrates that bind to the first enzyme but not to the second enzyme. In one embodiment, the system comprises a reporter protein that produces a detectable output. In one embodiment, the reporter protein generates a toxic product in the presence of a non-essential substrate. In one embodiment, the system further comprises an operon comprising a gene selected from the group consisting of a first inducible promoter of the first enzyme and a second inducible promoter of the second enzyme, wherein the second enzyme has an activity similar to the first enzyme.

In one embodiment, the present invention contemplates a method of using a microbial biosensor comprising an operon, wherein the operon comprises: a) providing a reporter gene and a sensor fusion protein gene; and b) expressing the sensor fusion protein having a post-translational modification and a reporter gene. In one embodiment, the expressed sensor fusion protein has a protein tyrosine phosphatase substrate domain and is capable of binding to the DNA binding sequence in the presence of at least one expressible sensor fusion protein that is a recognition domain (SH2) of the protein tyrosine phosphatase substrate domain attached to a phosphate molecule. In one embodiment, the operon further comprises gene fragments encoding: i) a first expressible sensor fusion protein as a protein tyrosine phosphatase substrate domain capable of attaching to the phosphate molecule, the first expressible sensor fusion protein being in operable combination with a DNA binding protein; and ii) a second expressible sensor fusion protein as a recognition domain (SH2) of the protein tyrosine phosphatase substrate domain when attached to a phosphate molecule, the second expressible sensor fusion protein being in operable combination with a subunit of RNA polymerase; and iii) individual expressible fragments, including but not limited to Src kinase protein, protein tyrosine phosphatase 1B (PTP1B) and conjugated to the transcriptionally active binding sequence of the DNA binding protein capable of binding to the sensor fusion protein, and the subunit of RNA polymerase being in operable combination with the reporter gene.

In one embodiment, the present invention contemplates a method of using a microbial biosensor, comprising: a) providing; i) an operon, wherein the operon comprises a reporter gene and a sensor fusion protein gene; ii) live bacteria; and iii) a test small molecule inhibitor of the protein tyrosine phosphatase; b) expressing the sensor fusion protein having a post-translational modification and a reporter gene; c) contacting the bacteria with the test small molecule; and d) determining whether the test small molecule is an inhibitor of the protein phosphatase by expression of the reporter gene. In one embodiment, the expressed sensor fusion protein has a protein tyrosine phosphatase substrate domain that is capable of binding to a DNA binding sequence in the presence of at least one expressible sensor fusion protein that is a recognition domain (SH2) of the protein tyrosine phosphatase substrate domain attached to a phosphate molecule. In one embodiment, the expressed sensor fusion protein has a protein tyrosine phosphatase 1B substrate domain that is capable of binding to the DNA binding sequence in the presence of at least one expressible sensor fusion protein that is a recognition domain (SH2) of the protein tyrosine phosphatase substrate domain attached to a phosphate molecule. In one embodiment, the operon further comprises a gene fragment encoding: i) the first expressible sensor fusion protein as the protein tyrosine phosphatase substrate domain capable of attaching to the phosphate molecule in an operable combination with DNA-binding; and ii) the second expressible sensor fusion protein as the recognition domain (SH2) of the protein tyrosine phosphatase substrate domain when attached to a phosphate molecule in an operable combination with a subunit of an RNA polymerase; and iii) a separate expressible fragment, which includes but is not limited to Src kinase protein, protein tyrosine phosphatase 1B (PTP1B) and conjugated to the transcriptionally active binding sequence of the DNA binding protein capable of binding to the sensor fusion protein, and the subunit of the RNA polymerase is in an operable combination with the reporter gene. In one embodiment, the biosensor further comprises an operon component expressing a second gene. In one embodiment, the biosensor further comprises an operon component expressing a second PTP different from the first PTP for identifying the inhibitor selective for one of the TPTases. In one embodiment, the test small molecule inhibitors include, but are not limited to, abietane-type diterpenes, abietic acid (AA), dihydroabietic acid, and structural variants thereof.

In one embodiment, the present invention contemplates a method of using a microbial biosensor, comprising: a) providing; i) an operon, wherein the operon comprises a reporter gene and a sensor fusion protein gene; ii) living bacteria; and iii) a test small molecule inhibitor of the protein tyrosine phosphatase; b) expressing the sensor fusion protein having a post-translational modification and a reporter gene; c) expressing the expressible sensor fusion protein in the bacteria; d) contacting the bacteria with the test small molecule; and e) determining whether the test small molecule is an inhibitor of the protein phosphatase by expression of the reporter gene. In one embodiment, the expressed sensor fusion protein has a protein tyrosine phosphatase substrate domain and is capable of binding to the DNA binding sequence in the presence of at least one expressible sensor fusion protein that is a recognition domain (SH2) of the protein tyrosine phosphatase substrate domain attached to a phosphate molecule. In one embodiment, the expressed sensor fusion protein has a protein tyrosine phosphatase 1B substrate domain and is capable of binding to the DNA binding sequence in the presence of at least one expressible sensor fusion protein that is a recognition domain (SH2) of the protein tyrosine phosphatase substrate domain attached to a phosphate molecule and a separately expressible fragment of the light-switchable protein tyrosine phosphatase 1B. In one embodiment, the operon includes gene segments encoding: i) the first expressible sensor fusion protein as the protein tyrosine phosphatase substrate domain capable of attaching to the phosphate molecule in operable combination with a DNA binding protein; ii) the second expressible sensor fusion protein as the recognition domain (SH2) of the protein tyrosine phosphatase substrate domain when attached to a phosphate molecule in operable combination with a subunit of an RNA polymerase; and iii) individual expressible segments including but not limited to Src kinase protein, protein tyrosine phosphatase 1B (PTP1B) and conjugated to the transcriptionally active binding sequence of the DNA binding protein capable of binding to the sensor fusion protein, and the subunit of the RNA polymerase in operable combination with the reporter gene.

In one embodiment, the present invention contemplates a method for providing a variant of a chemical structure for use as a potential therapeutic, comprising: a) providing; i) an E. coli bacterium comprising a metabolic terpenoid chemical structure production pathway that provides an altered chemical structure, wherein the metabolic pathway comprises a synthase, wherein the E. coli further comprises a microbial biosensor operon that detects PTP inhibition; and ii) a mutant synthase of an enzyme system; a) introducing a mutant synthase of the enzyme system; c) expressing the mutant synthase under conditions where the mutant synthase or enzyme system alters the chemical structure of the terpenoid chemical structure; and d) determining whether the altered chemical structure is an inhibitor of the PTP as a test inhibitor for use as a potential therapeutic. In one embodiment, the metabolic pathway comprises a synthase, including but not limited to a terpene synthase, a cytochrome P450, a halogenase, a methyltransferase, or a terpenoid functionalizing enzyme. In one embodiment, the terpenoid includes but is not limited to a terpenoid-type diterpene. In one embodiment, the terpenoid includes but is not limited to abietic acid.

In one embodiment, the present invention contemplates a fusion protein DNA construct comprising a protein phosphatase gene and a protein light switch gene conjugated within the phosphatase gene, wherein the protein phosphatase gene encodes a protein having a C-terminal domain and the protein light switch gene encodes a protein having an N-terminal alpha helical region, such that the C-terminal domain is conjugated to the N-terminal alpha helical region. In one embodiment, the construct further comprises an expression vector and a living cell. In one embodiment, the protein phosphatase is a protein tyrosine phosphatase. In one embodiment, the protein phosphatase is protein tyrosine phosphatase 1B (PTP1B). In one embodiment, the C-terminal domain encodes the α7 helix of PTP1B. In one embodiment, the construct encodes PTP1B _PS -A. In one embodiment, the construct encodes PTP1B _PS -B. In one embodiment, the protein phosphatase is a T-cell protein tyrosine phosphatase (TC-PTP). In one embodiment, the protein light switch is a light-oxygen-voltage (LOV) domain. In one embodiment, the protein light switch is the LOV2 domain of phototropin 1 from oats. In one embodiment, the LOV2 domain includes the A'a helix of LOV2. In one embodiment, the LOV2 has at least one mutation that results in an amino acid mutation. This is not meant to limit such mutations. In fact, mutations may include, but are not limited to, nucleotide substitutions, additions of nucleotides, and deletions of nucleotides from the gene. In one embodiment, the mutation is a nucleotide substitution. In one embodiment, the A'a helix of the LOV2 has a T406A mutation. In one embodiment, the protein photoswitch is a photosensitive protein. In one embodiment, the photosensitive protein is a bacterial photosensitive protein. In one embodiment, the bacterial photosensitive protein is a bacterial photosensitive protein 1 (BphP1) from Rhodopseudomonas palustris. In one embodiment, the protein photoswitch is a light-oxygen-voltage (LOV) domain with an artificial chromophore. In one embodiment, the protein photoswitch is a photosensitive protein with an artificial chromophore.

In one embodiment, the present invention contemplates a fusion protein comprising a protein phosphatase and a protein light switch conjugated within the phosphatase, wherein the protein phosphatase has a C-terminal domain and the protein light switch has an N-terminal alpha helical region, such that the C-terminal domain is conjugated to the N-terminal alpha helical region. In one embodiment, the fusion protein further comprises an expression vector and a living cell. In one embodiment, the protein phosphatase is a protein tyrosine phosphatase. In one embodiment, the protein phosphatase is a protein tyrosine phosphatase 1B (PTP1B). In one embodiment, the C-terminal domain encodes an α7 helix. In one embodiment, the fusion protein is PTP1B _PS -A. In one embodiment, the fusion protein is PTP1B _PS -B. In one embodiment, the protein phosphatase is a T-cell protein tyrosine phosphatase (TC-PTP). In one embodiment, the protein light switch is a light-oxygen-voltage (LOV) domain. In one embodiment, the protein light switch is the LOV2 domain of phototropin 1 from oats. In one embodiment, the LOV2 domain comprises the A'a helix of LOV2. In one embodiment, the A'a helix of the LOV2 has a T406A mutation. In one embodiment, the protein photoswitch is a light-oxygen-voltage (LOV) domain with an artificial chromophore. In one embodiment, the protein photoswitch is a photosensitive protein with an artificial chromophore. In one embodiment, the protein photoswitch is a photosensitive protein. In one embodiment, the photosensitive protein is a bacterial photosensitive protein. In one embodiment, the bacterial photosensitive protein is a bacterial photosensitive protein 1 (BphP1) from Rhodopseudomonas palustris. In one embodiment, the protein photoswitch is a light-oxygen-voltage (LOV) domain with an artificial chromophore. In one embodiment, the protein photoswitch is a photosensitive protein with an artificial chromophore.

In one embodiment, the present invention contemplates a method of using a fusion protein, comprising: a) providing; i) a fusion protein; ii) a protein phosphatase and iii) a living cell; and b) introducing the fusion protein into the living cell such that the illumination of the light switch changes a feature in the living cell. In one embodiment, the feature includes, but is not limited to, controlling cell movement, morphology, controlling cell signaling, and having a regulatory effect. In one embodiment, the regulatory effect includes, but is not limited to, inactivation, activation, reversible inactivation, and reversible activation. In one embodiment, the regulatory effect is dose-dependent. In one embodiment, the illumination is light in the range of 450-500nm. In one embodiment, the illumination is light in the range of 600-800nm. In one embodiment, the protein light switch undergoes a light-induced conformational change and the protein phosphatase has an allosterically regulated catalytic activity that is changed by the conformational change. In one embodiment, the change is enhanced or reduced. In one embodiment, the protein light switch is a light-oxygen-voltage (LOV) domain having an artificial chromophore. In one embodiment, the protein photoswitch is a phytochrome protein with an artificial chromophore. In one embodiment, the living cell is active. In one embodiment, the living cell is in vivo. In one embodiment, the method further comprises the step of controlling the activity of the cell in vivo.

In one embodiment, the present invention contemplates a method for detecting a small molecule modulator of a protein phosphatase, comprising: a) providing; i) a fusion protein comprising a protein phosphatase and a protein photoswitch; ii) a visual readout of phosphatase activity; iii) a light source, wherein the source is capable of emitting light radiation; iv) a living cell; and v) a small molecule test compound; b) expressing the fusion protein in the living cell; c) contacting the living cell with the small molecule test compound; d) irradiating the fusion protein within the cell with the light source; e) measuring a visual readout of changes in phosphatase activity for identifying the small molecule test compound as a modulator of the activity of the phosphatase; and f) using the modulating small molecule test compound for treating a patient exhibiting at least one symptom of a disease associated with the phosphatase. In one embodiment, the method further comprises identifying the small molecule test compound as an inhibitor of the activity of the phosphatase. In one embodiment, the method further comprises identifying the small molecule test compound as an activator of the activity of the phosphatase. In one embodiment, the disease includes, but is not limited to, diabetes, obesity, cancer, anxiety, autoimmunity, or neurodegenerative disease. In one embodiment, the protein photoswitch is a light-oxygen-voltage (LOV) domain with an artificial chromophore. In one embodiment, the protein photoswitch is a phytochrome protein with an artificial chromophore. In one embodiment, the method further provides a fluorescence-based biosensor and includes the step of introducing the fluorescence-based biosensor into the cell. In one embodiment, the method further includes the step of controlling the activity of the cell in vivo. In one embodiment, the visual readout of phosphatase activity is selected from the group consisting of a fluorescence-based biosensor, a change in cell morphology, and a change in cell motility.

In one embodiment, the present invention contemplates a light-operated switch protein tyrosine phosphatase construct comprising an N-terminal alpha helix of a protein light switch conjugated to a C-terminal allosteric domain region. In one embodiment, the protein tyrosine phosphatase is protein tyrosine phosphatase 1B (PTP1B). In one embodiment, the protein light switch is the LOV2 domain of phototropin 1 derived from oats (wild oats). In one embodiment, the enzyme construct further comprises an expression vector. In one embodiment, the present invention contemplates a biosensor for enzyme activity comprising: a) a substrate domain as described above; b) a substrate recognition domain; c) a first fluorescent protein; and d) a second fluorescent protein.

的均方跟偏差(RMSD)(如类似于PyMol函数对齐的函数定义的)。在一个实施方式中，所述蛋白磷酸酶的所述催化结构域与PTP1B的催化结构域具有至少34.1％序列同一性。在一个实施方式中，所述磷酸酶的所述催化结构域与PTP1B的催化结构域具有至少53.5％序列相似性。在一个实施方式中，所述蛋白激酶是蛋白酪氨酸激酶(PTK)。在一个实施方式中，所述蛋白激酶是PTK的催化结构域。在一个实施方式中，所述第一启动子是组成型启动子。在一个实施方式中，所述第二启动子是组成型启动子。在一个实施方式中，所述第一启动子是诱导型启动子。在一个实施方式中，所述第二启动子是诱导型启动子。在一个实施方式中，RNA聚合酶的所述结合位点包括部分第三启动子。在一个实施方式中，所述第一区域缺失分子伴侣的基因。在一个实施方式中，所述第一融合蛋白由连接能够将RNA聚合酶募集至DNA的蛋白质的底物识别结构域组成并且所述第二融合蛋白由连接至DNA结合蛋白质的底物结构域组成。在一个实施方式中，所述第一区域进一步含有包括与第一底物结构域不同的连接至不能够将RNA聚合酶募集至DNA的蛋白质的第二底物结构域的第三融合蛋白(即，“诱饵”)。在一个实施方式中，所述第三融合蛋白的所述底物结构域是所述激酶和所述磷酸酶二者的肽底物。在一个实施方式中，所述第三融合蛋白的所述底物结构域是所述激酶的肽底物，但是是所述磷酸酶的差的底物。在一个实施方式中，所述第一区域进一步含有第二蛋白磷酸酶的第六基因，其与第一蛋白磷酸酶不同并且其作用于所述第三融合蛋白的所述底物结构域。In one embodiment, the present invention provides a gene encoding system for detecting small molecules that regulate enzyme activity, which includes, a. a first region in an operable combination, which includes: i. a first promoter; ii. a first gene encoding a first fusion protein, the first fusion protein including a substrate recognition domain connected to a DNA binding protein; iii. a second gene encoding a second fusion protein, the second fusion protein including a substrate domain connected to a protein capable of recruiting RNA polymerase to DNA; iv. a second promoter; v. a third gene of a protein kinase; vi. a fourth gene of a molecular chaperone; vii. a fifth gene of a protein phosphatase; b. a second region in an operable combination, which includes: i. a first DNA sequence encoding an operator gene of the DNA binding protein; ii. a second DNA sequence encoding a binding site for RNA polymerase; and iii. one or more genes of interest (GOI). In one embodiment, the first promoter is Pro1. In one embodiment, the substrate recognition domain is a substrate homology 2 (SH2) domain from Homo sapiens. In one embodiment, the DNA binding protein is a 434 bacteriophage cI repressor. In one embodiment, the substrate domain is a peptide substrate for both the kinase and the phosphatase. In one embodiment, the second promoter is ProD. In one embodiment, the protein capable of recruiting RNA polymerase to DNA is the Ω subunit of RNA polymerase (i.e., RpoZ or RP _ω ). In one embodiment, the protein kinase is the Src kinase from Homo sapiens. In one embodiment, the molecular chaperone is CDC37 from Homo sapiens (i.e., Hsp90 co-chaperone). In one embodiment, the protein phosphatase is the protein tyrosine phosphatase 1B (PTP1B) from Homo sapiens. In one embodiment, the operator is the operator of 434 bacteriophage cI repressor. In one embodiment, the binding site of RNA polymerase is -35 to -10 region of lacZ promoter. In one embodiment, the gene of interest is SpecR, a gene that confers resistance to spectinomycin. In one embodiment, the gene of interest is LuxA and LuxB, two genes that produce luminescent output. In one embodiment, the gene of interest is a gene that confers resistance to antibiotics. In one embodiment, the protein phosphatase is PTPN6 from Homo sapiens. In one embodiment, the protein phosphatase is a protein tyrosine phosphatase (PTP). In one embodiment, the protein phosphatase is the catalytic domain of PTP. In one embodiment, an alignment of the X-ray crystal structures of (i) the catalytic domain of the protein phosphatase and (ii) the catalytic domain of PTP1B produces less than or equal to

The mean square root deviation (RMSD) of the alignment function (as defined by a function similar to the PyMol function alignment). In one embodiment, the catalytic domain of the protein phosphatase has at least 34.1% sequence identity with the catalytic domain of PTP1B. In one embodiment, the catalytic domain of the phosphatase has at least 53.5% sequence similarity with the catalytic domain of PTP1B. In one embodiment, the protein kinase is a protein tyrosine kinase (PTK). In one embodiment, the protein kinase is the catalytic domain of PTK. In one embodiment, the first promoter is a constitutive promoter. In one embodiment, the second promoter is a constitutive promoter. In one embodiment, the first promoter is an inducible promoter. In one embodiment, the second promoter is an inducible promoter. In one embodiment, the binding site of RNA polymerase includes part of the third promoter. In one embodiment, the first region lacks a gene for a molecular chaperone. In one embodiment, the first fusion protein consists of a substrate recognition domain connected to a protein capable of recruiting RNA polymerase to DNA and the second fusion protein consists of a substrate domain connected to a DNA binding protein. In one embodiment, the first region further contains a third fusion protein (i.e., "bait") comprising a second substrate domain that is different from the first substrate domain and is connected to a protein that is not capable of recruiting RNA polymerase to DNA. In one embodiment, the substrate domain of the third fusion protein is a peptide substrate for both the kinase and the phosphatase. In one embodiment, the substrate domain of the third fusion protein is a peptide substrate for the kinase, but is a poor substrate for the phosphatase. In one embodiment, the first region further contains a sixth gene for a second protein phosphatase that is different from the first protein phosphatase and acts on the substrate domain of the third fusion protein.

In one embodiment, the present invention provides a method of using: (i) a gene encoding system for detecting small molecules that modulate enzyme activity and (ii) a gene encoding pathway for terpenoid biosynthesis to identify and/or construct terpenoids that modulate enzyme activity, the method comprising, a. providing, i. a gene encoding system for detecting small molecules that modulate enzyme activity, comprising, 1. a first region in operable combination, comprising: a. a first promoter; b. a first gene encoding a first fusion protein, the first fusion protein comprising a substrate recognition domain linked to a DNA binding protein; c. a second gene encoding a second fusion protein, the second fusion protein comprising a substrate domain linked to a protein capable of recruiting RNA polymerase to DNA; d. a second promoter; e. a third gene for a protein kinase; f. . a fourth gene for a molecular chaperone; g. a fifth gene for a protein phosphatase; 2. a second region in operable combination, comprising: a. a first DNA sequence encoding an operator gene for the DNA binding protein; b. a second DNA sequence encoding a binding site for an RNA polymerase; c. one or more genes of interest (GOI); ii. a gene encoding a pathway for terpenoid biosynthesis, comprising: 1. a pathway for producing linear isoprenoid precursors; 2. a gene for a terpene synthase (TS); 3. a plurality of E. coli bacteria; b. transforming the bacteria with the following: (i) the gene encoding system for detecting small molecules and (ii) the gene encoding pathway for terpenoid biosynthesis, and allowing the transformed bacteria to replicate; c. observing the expression of the gene of interest through a measurable output. In one embodiment, the pathway for producing linear isoprenoid precursors produces farnesyl pyrophosphate (FPP). In one embodiment, the pathway for producing linear isoprenoid precursors is all or part of the mevalonate-dependent isoprenoid pathway of Saccharomyces cerevisiae. In one embodiment, the pathway to generate linear isoprenoid precursors is carried out through the plasmid pMBIS. In one embodiment, the gene of interest is SpecR, a gene that confers resistance to spectinomycin. In one embodiment, the TS gene is carried out on a separate plasmid (pTS) from the remaining terpenoid pathways. In one embodiment, the TS gene encodes amorphadiene synthase (ADS) from Artemisia annua. In one embodiment, the TS gene encodes gamma-humulene synthase (GHS) from Abies abies. In one embodiment, the TS gene encodes absinadiene synthase (ABS) from Abies abies, and the gene is operably combined with the gene for geranylgeranyl diphosphate synthase (GPPS). In one embodiment, the TS gene encodes taxene synthase (TXS) from Taxus brevis, and the gene is operably combined with the gene for GGPPS. In one embodiment, the method further comprises, d. extracting the terpenoid capable of the highest measurable output (e.g., growth at the highest concentration of spectinomycin); e. identifying the terpenoid; f. purifying the terpenoid. In one embodiment, the method further comprises, g. providing a mammalian cell culture, h. treating the cell culture with purified terpenoid, i. measuring a biochemical effect resulting from a change in the activity of a protein phosphatase or protein kinase. In one embodiment, the method further comprises, j. providing a purified enzyme target, k. measuring the modulatory effect of the purified terpenoid on the enzyme target, l. quantifying the modulatory effect (e.g., by calculating an IC ₅₀ ). In one embodiment, the TS gene has at least one mutation. In one embodiment, the TS gene is operably combined with a gene for an enzyme that functionalizes a terpenoid. In one embodiment, the TS gene is operably combined with a gene for a cytochrome P450. In one embodiment, the TS gene is operably combined with a gene for cytochrome P450BM3 from Bacillus megaterium. In one embodiment, the TS gene is operably combined with a gene for a halogenase. In one embodiment, the TS gene is an operable combination of a gene of a 6-halogenase (SttH) from streptomycin. In one embodiment, the TS gene is an operable combination of a gene of a vanadium halogen peroxidase (VHPO) from Acaryochloris marina. In one embodiment, the mammalian cell is a HepG2, Hela, Hek393t, MCF-7 and/or Cho-hIR cell. In one embodiment, the cell is a BT474, SKBR3 or MCF-7 and MDA-MB-231 cell. In one embodiment, the biochemical effect is insulin receptor phosphorylation, which can be measured by Western blot or enzyme-linked immunosorbent assay (ELISA). In one embodiment, the cell is a triple negative (TN) cell line. In one embodiment, the cell is a TN cell from the American Type Culture Collection (ATCC). In one embodiment, the cell is a TN cell from ATCC TCP-1002. In one embodiment, the biochemical effect is cell migration. In one embodiment, the biochemical effect is cell viability. In one embodiment, the biochemical effect is cell proliferation. In one embodiment, the protein phosphatase is PTP1B from Homo sapiens. In one embodiment, the protein kinase is Src kinase from Homo sapiens. In one embodiment, the gene of interest confers resistance to antibiotics. In one embodiment, the gene of interest is SacB, a gene that confers sensitivity to sucrose. In one embodiment, the gene of interest confers conditional toxicity (i.e., toxicity in the presence of an exogenously added molecule). In one embodiment, the genes of interest are SpecR and SacB. In one embodiment, the protein phosphatase is a wild-type enzyme. In one embodiment, the protein phosphatase has at least one mutation. In one embodiment, the protein phosphatase has at least one mutation that reduces its sensitivity to a small molecule that modulates the activity of the wild-type protein phosphatase. In one embodiment, the protein kinase is a wild-type enzyme. In one embodiment, the protein kinase has at least one mutation. In one embodiment, the protein kinase has at least one mutation that reduces its sensitivity to a small molecule that modulates the activity of the wild-type protein kinase. In one embodiment, the at least one of the terpenoids inhibits a protein phosphatase. In one embodiment, the at least one of the terpenoids inhibits PTP. In one embodiment, the at least one of the terpenoids inhibits PTP1B. In one embodiment, the at least one terpenoid activates a protein phosphatase. In one embodiment, the at least one terpenoid activates a PTP. In one embodiment, the at least one terpenoid activates protein tyrosine phosphatase non-receptor type 12 (PTPN12). In one embodiment, the at least one terpenoid inhibits a protein kinase. In one embodiment, the at least one terpenoid inhibits PTK. In one embodiment, the at least one terpenoid inhibits Src kinase. In one embodiment, the at least one terpenoid activates a protein kinase. In one embodiment, the at least one terpenoid activates a PTK. In one embodiment, the gene encoding system for detecting a small molecule further comprises both of the following: (i) a third fusion protein comprising a second substrate domain different from the first substrate domain linked to a protein that is unable to recruit RNA polymerase to DNA and (ii) a sixth gene for a second protein phosphatase different from the first protein phosphatase. In one embodiment, the gene encoding system for detecting small molecules further comprises both of the following: (i) a third fusion protein comprising a second substrate domain different from the first substrate domain linked to a protein that is incapable of recruiting RNA polymerase to DNA and (ii) a sixth gene for a second protein kinase different from the first protein kinase. In one embodiment, the gene encoding pathway for terpenoid biosynthesis alternatively comprises a library of pathways different from the identity of the TS gene, such that after transformation, most cells contain different TS genes (i.e., at least one gene with a different mutation). In one embodiment, the gene encoding pathway for terpenoid biosynthesis alternatively comprises a library of pathways different from the identity of genes for functionalized terpenoids (e.g., cytochrome P450 or halogenase) operably combined with SI genes, such that after transformation, most cells contain different genes for functionalized terpenoids (i.e., at least one gene with a different mutation). In one embodiment, the gene encoding pathway for terpenoid biosynthesis alternatively comprises a library of pathways in which the TS gene has been replaced by a component of a eukaryotic complementary DNA (cDNA) library, such that after transformation, most cells contain different genes in place of the TS gene. In one embodiment, the gene encoding pathway for terpenoid biosynthesis alternatively comprises a library of pathways in which the TS gene is accompanied by components of a eukaryotic complementary DNA (cDNA) library, such that after transformation, the majority of cells contain a different gene (e.g., a gene encoding an enzyme for terpenoid functionalization) in operable combination with the TS gene. In one embodiment, the gene encoding system for detecting small molecules alternatively comprises a library of such a system that is different in characteristics from the protein phosphatase gene, such that after transformation, the majority of cells contain different protein phosphatase genes (i.e., at least one gene that is mutated differently). In one embodiment, the gene encoding pathway for terpenoid biosynthesis generates terpenoids that modulate the activity of the wild-type form of the protein phosphatase, thereby enabling growth studies to isolate mutants of the protein phosphatase that are less sensitive to the modulatory effects of small molecules. In one embodiment, the gene encoding system for detecting small molecules alternatively comprises a library of such a system that is different in characteristics from the protein kinase gene, such that after transformation, the majority of cells contain separate protein kinase genes (i.e., at least one gene that is mutated differently). In one embodiment, the genetically encoded pathway for terpenoid biosynthesis produces terpenoids that modulate the activity of the wild-type form of the protein kinase, thereby enabling growth studies to isolate mutants of the protein kinase that are less sensitive to the modulatory effects of small molecules. In one embodiment, the at least one terpenoid modulates the activity of the wild-type form of the protein phosphatase, but not the mutant form of the protein phosphatase. In one embodiment, the at least one terpenoid modulates the activity of the first protein phosphatase, but not the second protein phosphatase. In one embodiment, the at least one terpenoid modulates the activity of the wild-type form of the protein kinase, but not the mutant form of the protein kinase. In one embodiment, the at least one terpenoid modulates the activity of the first protein kinase, but not the second protein kinase.

In one embodiment, the present invention provides an inhibitor detection operon, which includes, A: a first region in an operable combination under the control of a first promoter, which includes: i. a first DNA sequence encoding a first fusion protein including a substrate recognition homology 2 domain (SH2) and a repressor; ii. a second DNA sequence encoding a second fusion protein including a phosphate molecule binding domain of a substrate recognition domain, the substrate recognition domain and the ω subunit of RNA polymerase (RpoZ or RP _ω ); iii. a third DNA sequence encoding a cell division cycle 37 protein (CDC37); iv. a protein phosphatase; and B: a second region in an operable combination under the control of a second promoter, which includes: i. an operator gene including a repressor binding domain of the repressor, ii. a ribosome binding site (RB); and iii. a gene of interest (GOI). In one embodiment, the SH2 domain is a substrate recognition domain of the protein phosphatase. In one embodiment, the repressor is a 434 phage cI repressor. In one embodiment, the substrate recognition domain binds to the protein phosphatase. In one embodiment, the bait substrate domain is a Src kinase gene. In one embodiment, the operator is a 434cI operator. In one embodiment, the gene of interest encodes an antibiotic protein. In one embodiment, the protein phosphatase is a protein tyrosine phosphatase. In one embodiment, the first promoter is a constitutive promoter. In one embodiment, the second promoter is an inducible promoter.

In one embodiment, the present invention provides a method for detecting an operon using an inhibitor, comprising: a. providing, i. detecting an operon, comprising A: a first region in a combination operable under the control of a first promoter, comprising: 1. a first DNA sequence encoding a first fusion protein comprising a substrate recognition homology 2 domain (SH2) of a protein phosphatase and a repressor binding domain; 2. a phosphate molecule binding domain encoding a substrate recognition domain of a protein phosphatase, a substrate recognition domain of the protein phosphatase and an ω subunit of an RNA polymerase (RpoZ or _RPω )； 4. a third DNA sequence encoding a cell division cycle 37 (CDC37) protein； 5. a protein phosphatase； and B: a second region in combination operable under the control of a second promoter, comprising: 6. an operator comprising a repressor binding domain that binds the repressor, 7. a ribosome binding site (RB); and 8. a gene of interest (GOI); and ii. a mevalonate pathway operon with a deleted gene, such that the pathway operon does not contain at least one gene in the pathway that produces the terpenoid compound under the control of a third promoter comprising a second gene of interest for producing the terpenoid compound, i ii. a fourth DNA sequence under the control of a fourth promoter comprising the deleted gene from the mevalonate pathway operon and a third gene of interest; and iv. a plurality of E. coli bacteria, and b. transfecting the E. coli bacteria with the first operon for expressing the first gene of interest; c. transfecting the E. coli bacteria with the mevalonate pathway operon for expressing the first and second genes of interest; d. transfecting the E. coli bacteria with the fourth DNA sequence for expressing the first and second and third genes of interest; e. growing the cells, wherein the inhibitor terpenoid compound of protein phosphatase is produced by the cells. In one embodiment, the method further comprises the step of e. isolating the protein phosphatase inhibitor molecule and providing a mammalian cell culture for step f. treating the cell culture for reducing the activity of the protein phosphatase. In one embodiment, the method further comprises the step of providing an inducer compound for inducing the inducible promoter and contacting the bacteria with the compound. In one embodiment, wherein the method of reducing the activity of the protein phosphatase reduces the growth of the mammalian cells. In one embodiment, the protein phosphatase is human PTP1B. In one embodiment, the protein phosphatase is wild type. In one embodiment, the protein phosphatase has at least one mutation. In one embodiment, the deletion enzyme is a terpene synthase. In one embodiment, the terpene synthase is selected from the group consisting of amorphadiene synthase (ADS) and γ-humulene synthase (GHS). In one embodiment, the fourth DNA sequence further comprises geranylgeranyl diphosphate synthase (GPPS) and the deletion enzyme is selected from the group consisting of abietadiene synthase (ABS) and taxene synthase (TXS). In one embodiment, the terpene synthase is wild type. In one embodiment, the terpene synthase has at least one mutation. In one embodiment, the terpenoid compound is a structural variant of a terpenoid compound. In one embodiment, the gene of interest is an antibiotic gene. In one embodiment, each of the genes of interest is a different antibiotic gene.

In one embodiment, the gene encoding detection operator system comprises: Part A: a first region of DNA in an operable combination, comprising: a DNA region encoding a first promoter; a first gene encoding a first fusion protein, the first fusion protein comprising a substrate recognition domain connected to a DNA binding protein; a second gene encoding a second fusion protein, the second fusion protein comprising a substrate domain connected to a protein capable of recruiting RNA polymerase to DNA; a DNA region encoding a second promoter; a third gene of a protein kinase; a fourth gene of a molecular chaperone; a fifth gene of a protein phosphatase; Part B: a second region of DNA in an operable combination under the control of a second promoter, comprising: a first DNA sequence encoding an operator gene of the DNA binding protein; a second DNA sequence encoding a binding site for an RNA polymerase; and at least one gene of interest (GOI). In one embodiment, the substrate recognition domain is a substrate homology 2 (SH2) domain. In one embodiment, the DNA binding protein is a 434 phage cI repressor. In one embodiment, the substrate domain is a peptide substrate for both the kinase and the phosphatase. In one embodiment, the protein capable of recruiting RNA polymerase to DNA is the ω subunit (RPω) of RNA polymerase. In one embodiment, the gene of the kinase is the Src kinase gene. In one embodiment, the molecular chaperone is CDC37. In one embodiment, the molecular chaperone is the Hsp90 co-chaperone from Homo sapiens. In one embodiment, the operator is the 434 phage cI operator. In one embodiment, the gene of interest is a gene for antibiotic resistance. In one embodiment, the gene for antibiotic resistance produces an enzyme that allows bacteria to degrade antibiotic proteins. In one embodiment, the protein phosphatase is protein tyrosine phosphatase 1B. In one embodiment, the first and second promoters of part A are constitutive promoters. In one embodiment, the second promoter of part B is an inducible promoter.

In one embodiment, the present invention provides a method for using a gene-encoded detection operator system, which includes, a. providing, i. an inhibitor detection operator, which includes part A: a first region of DNA in an operable combination, which includes: 1. a DNA region encoding a first promoter; 2. a first gene encoding a first fusion protein, the first fusion protein including a substrate recognition domain linked to a DNA binding protein; 3. a second gene encoding a second fusion protein, the second fusion protein including a substrate domain linked to a protein capable of recruiting RNA polymerase to DNA; 4. a DNA region encoding a second promoter; 5. a third gene for a protein kinase; 6. a fourth gene for a molecular chaperone; 7. a fifth gene for a protein phosphatase; part B: a second region of DNA in an operable combination under the control of a second promoter, which includes: 8. a first DNA sequence of an operator gene encoding the DNA binding protein; 9. a second DNA sequence encoding a binding site for an RNA polymerase; and 10. at least one gene of interest (GOI). ii. a mevalonate-terpene pathway operon without a terpene synthase gene under the control of a fourth promoter comprising a second gene of interest producing terpenoid compounds, iii. a fourth DNA sequence under the control of a fifth promoter comprising the terpene synthase gene and a third gene of interest; and iv. a plurality of bacteria, and b. transfecting the bacteria with the inhibitor detection operon for expressing the first gene of interest; c. transfecting the bacteria with the mevalonate pathway operon for expressing the second gene of interest; d. transfecting the bacteria with the fourth DNA sequence for expressing the third gene of interest; e. growing the bacterial cells expressing the three genes of interest, wherein the inhibitor terpenoid compounds are produced by the bacterial cells inhibiting the protein phosphatase. In one embodiment, the method further comprises the step of e. isolating the protein phosphatase inhibitor molecule and providing a mammalian cell culture for step f. treating the cell culture for reducing the activity of the protein phosphatase. In one embodiment, wherein the method of reducing the activity of the protein phosphatase reduces the growth of the mammalian cells. In one embodiment, the protein phosphatase is human PTP1B. In one embodiment, the protein phosphatase is wild type. In one embodiment, the protein phosphatase has at least one mutation. In one embodiment, the mevalonate pathway operon includes genes for expressing mevalonate kinase (ERG12), phosphomevalonate kinase (ERG8), mevalonate pyrophosphate decarboxylase (MVD1), isopentenyl pyrophosphate isomerase (IDI gene) and farnesyl pyrophosphate (FPP) synthase (ispA). In one embodiment, the deleted enzyme is a terpene synthase. In one embodiment, the terpene synthase is selected from the group consisting of amorphadiene synthase (ADS) and γ-humulene synthase (GHS). In one embodiment, the fourth DNA sequence further includes geranylgeranyl diphosphate synthase (GPPS), and the terpene synthase is selected from the group consisting of abietadiene synthase (ABS) and taxene synthase (TXS). In one embodiment, the terpene synthase is wild type. In one embodiment, the terpene synthase has at least one mutation. In one embodiment, the terpenoid compound is a structural variant of a terpenoid compound. In one embodiment, the gene of interest is an antibiotic gene. In one embodiment, each of the genes of interest is a different antibiotic gene. In one embodiment, the method further provides an inducer compound for inducing the inducible promoter and the step of contacting the bacteria with the compound.

In one embodiment, the present invention provides a method using the following two: (i) a gene encoding system for detecting small molecules that regulate enzyme activity and (ii) a gene encoding pathway for polyketide biosynthesis to identify and/or construct polyketides that regulate enzyme activity, the method comprising providing a gene encoding system for detecting small molecules that regulate enzyme activity, which comprises a first region in an operable combination, which comprises: a first promoter; a first gene encoding a first fusion protein, the first fusion protein comprising a substrate recognition domain linked to a DNA binding protein; a second gene encoding a second fusion protein, the second fusion protein comprising a substrate domain linked to a protein capable of recruiting RNA polymerase to DNA; a second promoter; a third gene for a protein kinase; a fourth gene for a molecular chaperone; a fifth gene for a protein phosphatase; a second region in an operable combination, which comprises: a first DNA sequence encoding an operator gene for the DNA binding protein; a second DNA sequence encoding a binding site for an RNA polymerase; one or more genes of interest (GOI); a gene encoding pathway for polyketide biosynthesis, which comprises: a gene for a polyketide synthase; and a plurality of Escherichia coli bacteria. In one embodiment, the polyketide synthase is 6-deoxyerythrolide B synthase (DEBS). In one embodiment, the polyketide synthase (PKS) is a modular combination of different PKS components.

In one embodiment, the present invention provides a method using the following two: (i) a gene encoding system for detecting small molecules that modulate enzyme activity and (ii) a gene encoding pathway for polyketide biosynthesis to identify and/or construct an alkaloid that modulates enzyme activity, the method comprising, a. providing a gene encoding system for detecting small molecules that modulate enzyme activity, comprising, a first region in an operable combination, which comprises: a first promoter; a first gene encoding a first fusion protein, the first fusion protein comprising a substrate recognition domain connected to a DNA binding protein; a second gene encoding a second fusion protein, the second fusion protein comprising a substrate domain connected to a protein capable of recruiting RNA polymerase to DNA; a second promoter; a third gene for a protein kinase; a fourth gene for a molecular chaperone; a fifth gene for a protein phosphatase; a second region in an operable combination, which comprises: a first DNA sequence encoding an operator gene for the DNA binding protein; a second DNA sequence encoding a binding site for an RNA polymerase; one or more genes of interest (GOI); a gene encoding pathway for polyketide biosynthesis, which comprises, a pathway for alkaloid biosynthesis; a plurality of Escherichia coli bacteria. In one embodiment, the pathway for alkaloid biosynthesis is described herein.

In one embodiment, the present invention provides an engineered bacterial cell line comprising expression plasmid 1, plasmid 2, plasmid 3 and plasmid 4.

In one embodiment, the present invention provides a phosphatase inhibitor molecule produced by contacting a bacterium expressing plasmid 1 with an inducer molecule that induces a promoter expressing a terpenoid synthesis pathway operon in plasmid 2 and a terpene synthase in plasmid 3, wherein said plasmid 2 and plasmid 3 are co-expressed in said bacterium with plasmid 1. In one embodiment, said plasmid 2 and said plasmid 3 are under the control of an inducible promoter. In one embodiment, said bacterium is contacted with an inducible molecule for inducing said promoter.

In one embodiment, the present invention provides a bacterial strain that produces a phosphatase inhibitor molecule. In one embodiment, the inhibitor is a terpenoid molecule.

BRIEF DESCRIPTION OF THE DRAWINGS

The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawing(s) will be provided by the Office upon request and payment of the necessary fee.

1A-G illustrate an embodiment of developing a photoswitchable phosphatase, such as a PTP1B _PS , and show exemplary results.

FIG1A illustrates one embodiment of the design of the PTP1B _PS : light-induced unwinding of the A' α helix of LOV2 destabilizes the α7 helix of PTP1B, thus inhibiting catalysis. FIG1B illustrates one embodiment of the detailed description: In the competitive inhibition structure of PTP1B (orange), the α7 helix is stabilized and the WPD loop (black) adopts a closed, catalytically competent conformation. In the apolipoprotein (apo) structure (yellow), the α7 helix is disordered and the WPD loop (blue) adopts an open, inactive conformation. We attached the C-terminal α7 helix of PTP1B to the N-terminal A' α helix of LOV2 at the homologous intersection (1-7) to create a photoresponsive α7 helix destabilized chimera of LOV2. FIG1C shows an exemplary result of the optimization of one embodiment: construct 7 exhibits the largest dynamic range of the intersection variants; 7.1 has improved activity over 7, and 7.1(T406A) has improved dynamic range over 7.1. Figure 1D shows an exemplary analysis of the activity of PTP1B _PS on pNPP, indicating that light affects _kcat , but not _Km . Figure 1E shows that the dynamic range of PTP1B _PS is similar to that of substrates of different sizes. Figure 1F shows an exemplary illustration of two small molecules: p-nitrophenyl-phosphate (or pNPP) and 4-methylumbelliferyl phosphate (or 4MU) and a peptide domain from EGFR. Figure 1G shows exemplary activities of PTP1B-LOV2 chimeras that differ in (AD) crossover position and (E-E4) linker composition in the presence and absence of 455 nm light. Substrate: 4-methylumbelliferyl phosphate.

Figure 2A-J shows exemplary biophysical properties of the PTP1B _PS .

FIG2A shows exemplary mutations that (i) prevent the formation of cysteine adducts in LOV2 (C450M), (ii) destabilize the A'α and Jα helices in the case of LOV2 (I532E, I539E and ΔJα) or (iii) disrupt the allosteric network of PTP1B (Y152A/Y153A), reduce the photosensitivity by 7.1, except for I532E and C450M, reducing its specific activity. FIG2B shows exemplary exposure of PTP1B _PS to 455 nm light to reduce its α-helical content (CD _222nm ). FIG2C shows exemplary optical modulation of α-helical content (i.e., δ ₂₂₂ =CD _222-dark - _{CD 222-light} ) is necessary, but not sufficient, for optical modulation of catalytic activity. The dashed line represents δ ₂₂₂ of equimolar solutions of PTP1B _WT and LOV2 _WT . Figure 2D shows exemplary fluorescence of six tryptophan residues in the catalytic domain of PTP1B, which enables optical monitoring of its conformational state. Figures 2E-F show exemplary thermal recovery of (Figure 2E) α-helical content and (Figure 2F) tryptophan fluorescence of PTP1B _PS . Figure 2G shows that the exemplary kinetic constant for thermal resetting of α-helical content is greater than the exemplary kinetic constant for thermal resetting of tryptophan fluorescence, indicating that LOV2 resets faster than the PTP1B domain. The difference is minimal for the most photosensitive variant: 7.1 (T406A). Figure 2H shows an exemplary alignment of the crystal structures of PTP1B _PS (blue) and apolipoprotein PTP1B _WT (orange), indicating that LOV2 does not distort the structure of the catalytic domain. The LOV2 domain of PTP1B _PS may not be resolved; the flexible loop at the beginning of the α7 helix may cause LOV2 to adopt variable orientations in the lattice. The α6 and α7 helices of the inhibitory structure of PTP1B (yellow) are shown for reference. FIG2I shows an exemplary gap in the crystal structure of a PTP1B _PS that can accommodate LOV2. FIG2J shows an exemplary crystal of a PTP1B-LOV2 fusion that is green and becomes transparent when illuminated with 455 nm light; therefore, LOV2 is indeed present. The error bars in A, C, and G represent standard error (n>3). Note: The PTP1B _PS corresponds to construct 7.1 (T406A) from FIG1 .

3A-D illustrate exemplary fluorescence-based biosensors having PTP1B activity.

共振能量转移(FRET)导致CFP荧光减少和YFP荧光增加；当传感器是以其磷酸化状态时，FRET的不存在导致相反的作用。图3B显示了供体荧光(CFP)与接纳体荧光(YPet)的比例的示例性增加，证实存在Src激酶(即，酪氨酸激酶)。当另外添加(i)EDTA，其螯合Src的金属辅助因子或(ii)PTP1B，其使底物结构域去磷酸化时，该增加不发生。图3C显示了作为A的FRET传感器的另一种类的一个实施方式；该实施方式使用mClover3和mRuby3。这些蛋白质的激发和发射波长使得它们与基于LOV2的成像实验相容。图3D显示了用来自C的传感器示例性重复来自B的实验。FIG3A shows one embodiment of a sensor for PTP1B activity. The sensor consists of a kinase substrate domain, a short flexible linker, and a phosphorylation recognition domain sandwiched between two fluorescent proteins (e.g., cyan fluorescent protein and yellow fluorescent protein). When the sensor is in its unphosphorylated state, the phosphorylation between the two fluorophores is

Resonance energy transfer (FRET) results in a decrease in CFP fluorescence and an increase in YFP fluorescence; the absence of FRET results in the opposite effect when the sensor is in its phosphorylated state. FIG. 3B shows an exemplary increase in the ratio of donor fluorescence (CFP) to acceptor fluorescence (YPet), confirming the presence of Src kinase (i.e., tyrosine kinase). This increase does not occur when (i) EDTA, which chelates metal cofactors of Src, or (ii) PTP1B, which dephosphorylates the substrate domain, is additionally added. FIG. 3C shows an embodiment of another class of FRET sensors as A; this embodiment uses mClover3 and mRuby3. The excitation and emission wavelengths of these proteins make them compatible with LOV2-based imaging experiments. FIG. 3D shows an exemplary repetition of the experiment from B with the sensor from C.

4A-H show exemplary evidence of phosphatase activity in living cells using photoconstructs and fluorescent tags.

Figures 4A-C show embodiments of three constructs expressed in Cos-7 cells: (Figure 4A) GFP-PTP1B _PS , (Figure 4B) GFP-PTP1B _PS -A, and (Figure 4C) GFP-PTP1B _PS -B. Here, GFP-PTP1B _PS is a fusion of green fluorescent protein (GFP) and the N-terminus of 7.1 (T406A) from Figures 1B-C (without a histidine tag); GFP-PTP1B _PS -A is a fusion of GFP-PTP1B _PS and the C-terminal domain of full-length PTP1B; and GFP-PTP1B _PS -B is a fusion of GFP-PTP1B _PS and the C-terminal endoplasmic reticulum (ER) anchor of full-length PTP1B (see below). GFP-PTP1B _PS is localized in the cytosol and nucleus, while GFP-PTP1B _PS -A and GFP-PTP1B _PS -B are localized in the ER. Figure 4D-H shows exemplary results of cell-based studies of PTP1B _PS . We transformed Cos-7 cells with plasmids containing (i) FRET sensors from Figures 3C-3D and (ii) PTP1B _PS or PTP1B _PS / C450M (light-insensitive mutants). In this experiment, we illuminated individual cells with 447nm light and immediately imaged them with 561nm light. Changes in the light regulation of the FRET ratio (as defined in Figure 3) allow us to detect changes in the light regulation of PTP1B activity. Figure 4D-E shows exemplary Cos-7 cells transformed with PTP1B _PS at two time points: (Figure 4D) immediately after excitation with 447nm light and (Figure 4E) 1min later. A slight increase in the FRET ratio (dark green to light green) confirms the photoactivation of PTP1B. (FG). Cos-7 cells were transformed with PTP1B _PS (C450M) at two time points: (FIG. 4F) immediately after excitation with 447 nm light and (G) 1 min later. The absence of detectable changes in FRET ratios indicates that the changes observed in DE are caused by light-induced changes in PTP1B activity. FIG. 4H shows exemplary mean fractional changes in FRET ratios observed in the nucleus (nuc) and cytosol (cyt) after 1 min and 2.67 min. The changes in PTP1B _PS are higher than those in the light-insensitive mutant PTP1B _PS (C450M). Error bars indicate standard errors.

5A-C illustrate embodiments of drug discovery.

Fig. 5A shows an exemplary use of a phosphatase, i.e., a drug target (upper left figure of PTP1B), for identifying a synthetase (lower right figure), wherein the enzyme is then used to provide an inhibitor or modulator molecule of the phosphatase, thus showing a general framework for using an enzyme to construct an inhibitor of a selected protein target. Fig. 5B shows an exemplary analysis of structural relationships between binding pockets. The matrix compares each characteristic (e.g., volume) between binding pocket 1 and all other binding pockets (2 to n) that can be functionalized (e.g., P450) or bound to (e.g., PTP1B) the ligand synthesized in pocket 1. Fig. 5C shows an exemplary comparison of the ability of binding pockets to bind intermediates in a biosynthetic pathway.

Figure 6 illustrates an overlay of PTP1B (PDB entries 1t4j and 2f71, respectively) showing allosterically inhibited (green) and competitively inhibited (orange) structures of PTP1B, demonstrating activity-regulated conformational changes: Unwinding of the α7 helix of LOV2 (blue) causes its catalytically essential WPD loop (right) to adopt an open, catalytically compromised conformation. Competitive (red) and allosteric (yellow) inhibitors highlight the active site and allosteric site, respectively.

Figures 7A-B show exemplary analyses of binding affinity. Figure 7A shows an embodiment of two binding partners of PTP1B: LMO4 and Stat3. Figure 7B shows an exemplary binding isotherm based on binding-induced changes in tryptophan fluorescence of PTP1B (ligand is TCS 401, a competitive inhibitor).

Figures 8A-B illustrate exemplary structural alignments of PTP1B (light blue) and STEP (orange), Figure 8A, which have only 31% sequence identity, showing significant structural similarity. Figure 8B illustrates an exemplary structure of PTK6. Both STEP and PTK6 possess C-terminal α-helices that are compatible with the drive of the N-terminal helix of LOV2 (i.e., the light modulating framework is similar to that depicted in Figure 1).

Figure 9A-B illustrates an exemplary framework for constructing an enzyme regulated by red light.We attached the C-terminal α-helix of PTP1B to the N-terminal α-helix of BphP1.

Figures 10A-B illustrate exemplary operons for screening light-operated switch variants of PTP1B. Figure 10A shows an exemplary illustration that PTP1B in its active state (here, far-red state) dephosphorylates the substrate domain, preventing substrate-SH2 association, and thus preventing transcription. Figure 10B shows an exemplary illustration that the substrate domain phosphorylated in its inactive state (here, red state) binds SH2, allowing transcription of a gene for antibiotic resistance.

Figures 11A-B illustrate an exemplary strategy for the evolution of photoswitchable proteins. Figure 11A illustrates that we will compare the growth of colonies on replicate plates exposed to red and infrared light, and select colonies that exhibit differential growth. Figure 11B illustrates that we will further characterize the photosensitivity of hothit in liquid culture.

Figures 12A-B illustrate exemplary FRET-based sensors developed to measure intracellular phosphatase or kinase activity. Binding of substrate to the SH2 domain enhances (Figure 12A) or reduces (Figure 12B) FRET, depending on the architecture.

An animation of the imaging experiment is shown in Figure 13. We will inactivate PTP1B PS within subcellular compartments (1-10 μm) containing varying amounts of plasma membrane, ER, and cytosol, and we will use fluorescence lifetime imaging to examine the phosphorylation state of our FRET-based sensor (from Figure 12) in whole cells.

14A-D illustrate exemplary starting points for guided drug design and discovery.

FIG. 14A illustrates abietic acid. FIG. 14B shows inhibition of PTP1B by abietic acid at concentrations of 0-400 uM (dark to light). Analysis of different fits indicates non-competitive or mixed inhibition. FIG. 14C illustrates abietic acid docked in the allosteric site of PTP1B (green). As we have shown that abietic acid binds to the active site of PTP1B. The inset highlights the active site in black. FIG. 14D shows an exemplary X-ray crystal structure of a known allosteric inhibitor (blue).

Figures 15A-D illustrate an exemplary Figure 15A of a terpenoid synthesis pathway (mevalonate can be synthesized or added to the media by pMevT). Figure 15B shows an exemplary abietadiene titer produced by E. coli DH5a transformed with a plasmid from A (without P450). Figures 15C-D show exemplary GC-MS analysis of the following products of a strain producing abietadiene in Figure 15C) and a strain producing abietic acid in Figure 15D: (1) abietadiene, (2) levorotatory pinamaradiene, and (3) abietic acid (10,000 ion counts for Figure 15C and 1,000 ion counts for Figure 15D). Note: Avoiding protein overexpression in E. coli DH5a is common in metabolic engineering* ⁴ .

Figures 16A-C illustrate exemplary terpenoids showing differences in stereochemistry, shape, size, and chemical functionality. Figure 16A illustrates clockwise abietic acid (1), neoabietic acid (2), levopimaric acid (3), dihydroabietic acid (4). Figure 16B shows exemplary initial rates of PTP1B to 10 mM p-NP phosphate in the presence of 200 uM inhibitor. No inhibitor (Figure 16C). Error bars = standard error (n>5).

Figures 17A-C show results from an exemplary study. Figure 17A N-HSQC spectra of PTP1B (red) and PTP1B bound to abietic acid (blue). Inset: Crystal structure of PTP1B. Figures 17B-C show exemplary tryptophan (W) fluorescence of PTP1B in the presence of culture extracts of the control (ABS _X ), abietic diene (ABS) producing strains and abietic acid (ABS/BM3) of Figure 17B and (Figure 17C) various concentrations of abietic acid and 25uM of a known allosteric inhibitor (BBR). Error bars represent standard errors (n>5).

Figures 18A-B illustrate exemplary terpenoids that differ in stereochemistry Figure 18A and shape Figure 18B. Inset: Residues targeted for mutagenesis in the class I position of the ABS.

Figures 19A-E illustrate exemplary terpenoids Figure 19A carboxylated, Figure 19B hydroxylated, and Figure 19C halogenated diterpenes. Figures 19D-E show exemplary residues for targeted mutagenesis in Figure 19D P450 BM3 and Figure 19E SttH.

An exemplary WaterMap analysis of UPPS is illustrated in Figure 20. The color of the water molecules corresponds to the free energy relative to the bulk water.

21A-E illustrate an exemplary high throughput screen for PTP1B inhibitors.

FIG. 21A Growth coupling (i.e., selection; strategy 1). FIG. 21B) FRET sensor for PTP1B activity (strategy 2). FIG. 21C) FRET sensor and FIG. 21D) tryptophan fluorescence for PTP1B conformational changes (strategies 3 and 4). FIG. 21E, results similar to those for the operon shown in FIG. 21A, with Lux substituted for Amp. Error bars = SD (n≥3).

Figures 22A-D illustrate exemplary inhibition of PTP1 B. Error bars in Figure 22C represent SE (n > 3 independent reactions).

FIG. 22A shows an exemplary alignment of the backbone of PTP1B in competitive inhibition (yellow and orange, PDB entry 2F71) and allosteric inhibition (grey and black, PDB entry 1T4J) poses. Binding of substrate and competitive inhibitor to the active site causes the WPD loop to adopt a closed (orange) conformation, which stabilizes the C-terminal α7 helix through an allosteric network; the helix is indecomposable in allosterically inhibited, noncompetitively inhibited, and uninhibited structures, which exhibit a WPD-open conformation (black). FIG. 22B shows an exemplary illustration of the chemical structure of abietic acid (AA). FIG. 22C shows an exemplary initial rate of PTP1B-catalyzed hydrolysis of pNPP in the presence of increasing concentrations of AA. The line shows a fit to a model of mixed inhibition. FIG. 22D illustrates the model in which the inhibitor (I) binds to the enzyme (E) and the enzyme-substrate complex (ES) with different affinities.

23A-C illustrate exemplary NMR analysis of PTP1B-AA association.

Figure 23A shows an exemplary weighted difference in chemical shifts (Δ) between 1H-15N-HSQC spectra collected in the absence and presence of AA (10:1 PTP1B:AA). The red dashed line depicts the threshold for Δ values greater than two standard deviations (σ) above the mean; the gray bar marks the residues whose chemical shifts are expanded beyond the recognition. Figure 23B illustrates an exemplary crystal structure of PTP1B (PDB entry 3A5J, gray) highlighting the position of the specified residues (blue); overlays of inhibitors in the allosteric site (PDB entry 1T4J, green) and the active site (PDB entry 3EB1, yellow) are used for reference. Residues with significant CSPs (i.e., Δ>Δ mean + 2σ) are distributed throughout the protein (red) and, except for two residues in the WPD loop, outside the known binding site. Figure 23C illustrates exemplary details of the active site (top) and the known allosteric site (bottom), with inhibitors from (Figure 23B) overlays.

Figure 24A-C illustrates exemplary mutational analysis of the AA binding site.

FIG. 24A illustrates an exemplary crystal structure of PTP1B (gray, PDB entry 3A5J), which shows the locations of mutations introduced at the following five sites: active site (red), allosteric site (green), site 1 (orange), site 2 (yellow), and L11 loop (blue). The binding configurations covering BBR (allosteric site, PDB entry 1T4J) and TCS401 (active site, PDB entry 1C83) are used for reference. FIG. 24B illustrates exemplary disruptive mutations introduced at each site. Mutations are designed to change the size and/or polarity of the targeted residues. Mutations represented as "YAYA" (Y152A/Y153A), which were identified in previous studies, weakened the allosteric communication between the C-terminus and the WPD loop. FIG. 24C illustrates exemplary fractional changes in inhibition (F in Eq. 1) caused by mutations from (B). Five mutations distributed throughout the protein reduce inhibition by AA and TCS401, but have negligible effects on inhibition by BBR. The similar effects of most mutations on AA and TCS401 suggest that both inhibitors bind to the active site. Error bars represent SE (propagation from n≥9 independent measurements for each V in Eq. 1).

Figure 25A-D illustrates exemplary computational analysis of AA binding.

Figure 25A-B illustrates the exemplary results of molecular dynamics simulation: (A) the skeleton trace (trace) of PTP1B without AA and Figure 25B illustrate the exemplary amino acid (AA)-binding state. The thickness of the trace indicates the amplitude and direction of the local motion (method). The combination of AA increases the flexibility of WPD, E and L10 rings. The WPD and L10 rings contain residues (red) with significant CSP, indicating the consistency between the results of MD and NMR analysis. Figure 25C illustrates the exemplary representative binding conformation of AA (green). In binding to the active site, AA (i) forms hydrogen bonds with R221 that weakens the combination between R221 and E115 and (ii) prevents the formation of hydrogen bonds (red) between W179 and R221 formed when the WPD ring is closed. Two effects enhance the conformational dynamics of the WPD ring. Figure 25D shows exemplary results of docking calculations, consistent with mixed-type inhibition: binding of AA prevents WPD ring closure and disruption, but does not exclude, binding of pNPP (blue spheres).

Figures 26A-B illustrate exemplary terpenoids showing differences in stereochemistry, shape, size, and chemical functionality. Figure 26A) Structural analogs of abietic acid (AA): abietic acid (CA), isopimaric acid (IA), dehydroabietic acid (DeAA), and dihydroabietic acid (DiAA). Figure 26B) Differences in saturation produce significant differences in potency (i.e., IC50), but not selectivity. Error bars represent 95% confidence intervals.

27A-C Analysis of pathology-associated mutations.

Figure 27A illustrates an exemplary bar graph of a mutation characterized by kinetics. All mutations close to (<4A) five or more network residues are "influential" (i.e., they change _kcat or _KM >50% or have a detectable effect on inhibition); on the contrary, non-consequential mutations have fewer neighboring network residues. Figure 27B illustrates an exemplary crystal structure of PTP IB (gray, PDBruk 3A5J), highlighting the position of influential mutations on network residues; colors indicate whether they are introduced into biophysical studies or found in diseases. Figure 27C illustrates two exemplary cumulative distribution functions, describing the number of randomly selected network residues close to (i) mutations identified in the disease and (ii) sites. The two distributions are indistinguishable from each other (P<0.05), indicating that disease-related mutations do not preferably occur near the allosteric network.

Figures 28A-D illustrate and show exemplary data using a gene operon that links PTP activity to the output of a gene of interest (GOI).

Figure 28A shows an embodiment of operon A. Example of an operon. S, tyrosine substrate; P, phosphate group; cI, 434 phage cI repressor; RpoZ and _RPω , ω subunits of RNA polymerase; cI OP, binding sequence of 434 phage cI repressor; and RB, binding site of RNA polymerase (RNAP). Phosphorylation of the tyrosine substrate (by c-Src kinase) leads to binding of the substrate- _RPω fusion to the SH2-cI fusion; this binding event then positions the RNA polymerase to the RB, triggering transcription of the GOI. PTP1B dephosphorylates the substrate domain, preventing the association of the substrate- _RPω fusion and SH2-cI, thereby stopping transcription of the GOI. Then, inactivation of PTP1B re-enables transcription of the GOI. FIG. 28B illustrates an embodiment of a proposed medium-throughput screening for membrane-permeable inhibitors: strains of E. coli are transformed with operons and grown in the presence of small molecules; small molecule inhibitors of PTP1B regulate transcription of GOIs (e.g., genes for luminescence, fluorescence, or antibiotic resistance) in a dose-dependent manner. The bar graph shows the predicted trend of the data. FIG. 28C shows an embodiment of operon B. The operon can be screened for selective inhibitors. The operon includes operon A, which has (i) a second substrate-protein fusion (red), a "bait" that can bind to the SH2-cI fusion, but does not specifically bind to DNA or RNA polymerase, and (ii) a second PTP (e.g., TC-PTP) that is active on the substrate domain of the bait. Because the complex between the bait and SH2-cI does not trigger transcription, the bait inhibits transcription by competing with the cI-substrate for binding sites. Accordingly, molecules that inhibit PTP1B but not TC-PTP (which dephosphorylates the bait)—i.e., selective inhibitors—cause maximum transcriptional activation. On the contrary, molecules that inhibit both enzymes result in less activation. Figure 28D shows an embodiment of operon C. The operon enables screening of light-operated switchable enzymes. The operon includes a version of operon A in which PTP1B has been replaced with a light-operated switchable version of PTP1B. In this case, transcription of the GOI is different (e.g., higher or lower) under different light sources. In the example shown, light inhibits the activity of the PTP1B-LOV2 chimera, therefore, enhancing transcription of the GOI.

Figure 29A-B shows exemplary preliminary results of phosphorylation-dependent expression of GOI. Figure 29A shows an embodiment of operon A, wherein GOI is bacterial luciferase (LuxAB). PTP1B suppresses luminescence (i.e., reduces the transcription of GOI), and the catalytic non-activated version of PTP1B (the analogue of the inhibition version of PTP1B) enhances luminescence. Figure 29B shows an embodiment of operon A, wherein GOI is a gene (SpecR) to spectinomycin resistance. PTP1B suppresses spectinomycin growth, and the catalytic non-activated version of PTP1B (the analogue of the inhibition version of PTP1B) enhances growth. Used herein is MidT substrate.

Figure 30A-B illustrates operon optimization.

Figure 30A shows an embodiment of an operon from A, wherein GOI is bacterial luciferase (LuxAB), PTP1B is missing, and substrates are peptides from Kras, midT, ShcA or EGFR. Although all substrates can be phosphorylated by Src kinase, only two substrates bind tightly enough to the SH2 domain to enable significant luminescence above background (0% arabinose). Figure 30B shows an embodiment of an operon from A (here, contained on a single plasmid), wherein GOI is bacterial luciferase (LuxAB) and PTP1B is missing. The Y/F mutation (blue) on the substrate domain prevents it from being phosphorylated. The RBS site triggers the expression of Src kinase.

Figure 31A-D illustrates the use of operons.

FIG. 31A illustrates an exemplary concept for screening for microbially synthesizable inhibitors of PTP1B. When transformed with an embodiment of operon A (or operon B), cells capable of synthesizing PTP1B-inhibiting metabolites will produce different GOI outputs than cells that do not produce such metabolites. Because abietane-type diterpenes can (i) inhibit PTP1B and (ii) be synthesized in E. coli, we believe that strains of E. coli containing both operon A and pathways for constructing abietane-type diterpenes can be "evolved" to construct inhibitors of PTP1B. Here, the GOI can be a gene for luminescence or fluorescence (low flux) or antibiotic resistance (high flux). FIG. 31B illustrates an exemplary concept for screening for light-operated switchable enzymes. Fusions of PTP1B with LOV2 or BphP1 are considered (here, the highlighted helix shows the N-terminal connection point on these two proteins). For this example, irradiation of PTP1B-LOV2 with 455 nm light reduced its activity; irradiation of the PTP1B-BphP1 fusion with 650 nm light reduced its activity, while irradiation of the PTP1B-BphP1 fusion with 750 nm light enhanced its activity. When transformed with Operon C (which would contain one of these fusions), cells would produce different GOI outputs under different irradiation conditions. FIG. 31C illustrates an exemplary concept for screening for selective mutants of an enzyme. When transformed with a version of Operon B in which (i) PTP1B is also active on the bait and (ii) the second PTP (TC-PTP in our example) is missing, cells containing a mutant of PTP1B will most efficiently transcribe the GOI when PTP1B is only active on the bait substrate. FIG. 31D illustrates an exemplary concept for screening for selective substrates. When transformed with a version of operon B in which (i) the bait is deleted, (ii) the first enzyme (PTP1B in our example) is under an inducible promoter, (iii) the second PTP (TC-PTP in our example) is under a second inducible promoter, and (iv) the GOI includes genes for antibiotic resistance and a gene that produces a toxic product in the presence of a non-essential substrate, cells containing a mutated substrate domain will grow under both condition 1 (inducer of PTP1B and non-essential substrate) and condition 2 (inducer of TC-PTP) when they bind to PTP1B but not to TC-PTP.

32A-B present exemplary evidence for an evolutionarily conserved allosteric network.

Figure 32A relates to exemplary results of statistical coupling analysis. The orange and blue clusters represent two groups of interconnected residues, called "sectors", which exhibit strong intra-group correlations in the non-random distribution of amino acids. Highlight allosteric sites (green inhibitors, PDB entry 1T4J), WPD loops (purple balls) and active sites (red inhibitors, 3EB1) for reference. Figure 32B relates to an exemplary analysis of crosstalk between pockets of PTP1B modeled by MD simulation. Pockets are represented as balls, colored according to their persistence along the MD trajectory; the size of each ball indicates its average volume in the MD simulation. The thickness of the connection is proportional to the frequency of merging and splitting events (i.e., communication) in the middle of the pocket. The two independent groups of interconnected pockets are closely linked to the partitions identified in SCA, therefore, indicating that the two partitions represent different domains of the evolutionarily conserved allosteric network. In the PTP1B-LOV2 fusion of Figure 1, LOV2 regulates the activity of PTP1B by utilizing the allosteric network defined by partition A. Thus, the identification of partition A by statistical coupling analysis of the PTP family indicates that the architecture of light control described in Figure 1 is broadly applicable to all protein tyrosine phosphatases.

Figures 33A-E illustrate embodiments of a genetically encoded system for ligating enzyme activity to expression of a gene of interest (GOI). Error bars in Figures 33B-E represent standard deviation, n=3 biological replicates.

FIG33A illustrates an embodiment of a bacterial two-hybrid system for detecting phosphorylation-dependent protein-protein interactions. Components include (i) a substrate domain fused to the ω subunit of RNA polymerase (yellow), (ii) an SH2 domain fused to the 434 phage cI repressor (light blue), (iii) an operator of 434cI (dark green), (iv) a binding site for RNA polymerase (purple), (v) Src kinase, and (vi) PTP1B. Src-catalyzed phosphorylation of the substrate domain enables the substrate-SH2 interaction to activate transcription of the gene of interest (GOI, black). PTP1B-catalyzed dephosphorylation of the substrate domain prevents the interaction; inhibition of PTP1B re-enables it. FIG33B relates to an embodiment of the two-hybrid system from FIG33A that (i) lacks PTP1B and (ii) contains luxAB as GOI. We used inducible plasmids to increase expression of specific components; overexpression of Src enhanced luminescence. FIG33C relates to an embodiment of the two-hybrid system from FIG33A that (i) lacks both PTP1B and Src and (ii) includes a "superbinder" SH2 domain (SH2*, i.e., an SH2 domain with mutations that enhance its affinity for phosphopeptides), a variable substrate domain, and LuxAB as a GOI. We used an inducible plasmid to increase the expression of Src; luminescence increased most prominently for p130cas and MidT, indicating that Src acts on both substrate domains. FIG33D relates to an embodiment of the two-hybrid system from FIG33C with one of two substrates: p130cas or MidT. We used a second plasmid to overexpress (i) Src and PTP1B or (ii) Src and an inactive variant of PTP1B (C215S). The difference in luminescence between systems containing PTP1B or PTP1B(C215S) was greatest for MidT, indicating that PTP1B acts on this substrate. Right: An optimized version of the two-hybrid system (bb030 as the RBS of PTP1B) is presented for reference. FIG. 33E shows the results of an exemplary growth-coupled assay using an optimized B2H comprising SH2*, midT substrate, an optimized promoter and ribosome binding site (bb034 of PTP1B) and SpecR as the GOI. The system is illustrated at the top of the figure. The exemplary growth results show that inactivation of PTP1B enables strains of E. coli to carry the system to survive high concentrations of spectinomycin (>250 μg/ml).

FIG. 34 illustrates an exemplary experiment for optimizing the B2H system depicted in FIG. 33 .

FIG. 35 illustrates an exemplary experiment for optimizing the B2H system depicted in FIG. 33 for use in a growth-coupled assay.

36A-C depict exemplary metabolic pathways for the biosynthesis of terpenoids.

Figure 36A depicts plasmid loading pathways for terpenoid biosynthesis: (i) pMBIS, which carries the mevalonate-dependent isoprenoid pathway of Saccharomyces cerevisiae, converting mevalonate to isopentanyl pyrophosphate (IPP) and farnesyl pyrophosphate (FPP). (ii) pTS, which encodes a terpene synthase (TS) and, if necessary, a geranylgeranyl diphosphate synthase (GPPS), converting IPP and FPP to sesquiterpenes and/or diterpenes.

36B depicts exemplary terpene synthases: amorphadiene synthase (ADS) from Artemisia annua, gamma-humulene synthase (GHS) from Abies selengensis, abietadiene synthase (ABS) from Abies selengensis, and taxene synthase (TXS) from Taxus brevifolia.

Figure 36C shows the results of an exemplary growth-coupled assay of an E. coli strain containing both (i) an embodiment of an optimized bacterial two-hybrid (B2H) system (i.e., the B2H system from Figure 33E) and (ii) an embodiment of a pathway for terpenoid biosynthesis (i.e., the pathway from Figure 35A).

Figures 37A-C provide exemplary analysis of the inhibitory effects of terpenoids produced by different strains of E. coli.

Figure 37A depicts the results of our analysis of the inhibitory effects of DMSO containing (i) no inhibitor and (ii) extracted compounds from the culture fermentation broth of strains containing ADS. Figure 37B depicts the results of our analysis of the inhibitory effects of DMSO containing (i) extracted compounds from the culture fermentation broth of strains containing GHS (gHUM) or (ii) extracted compounds from the culture fermentation broth of strains including the L450Y mutant of GHS. Figure 37C depicts the results of our analysis of the inhibitory effects of DMSO containing (i) no inhibitor, (ii) extracted compounds from the culture fermentation broth of strains containing ABS, (iii) extracted compounds from the culture fermentation broth of strains containing TXS, and (iv) extracted compounds from the culture fermentation broth of strains containing catalytically inactive variants of ABS.

FIG. 38 shows an exemplary analysis of the product profile of GHS mutants capable of growth in the presence of spectinomycin.

FIG. 39 shows analysis of an exemplary B2H system linking inhibition of other PTPs to cell survival.

Figures 40A-E depict exemplary embodiments of genetically encoded systems that link enzyme activity to expression of a gene of interest, and those embodiments are applied to (i) prediction of resistance mutations, (ii) construction of inhibitors to counteract resistance mutations, and (ii) evolution of kinase inhibitors.

FIG. 40A depicts an exemplary first step in examining potential resistance mutations. By evolving metabolic pathways to produce molecules that inhibit known drug targets (e.g., PTP1B); these molecules will allow expression of a gene of interest (GOI) that confers survival in the presence of selective pressure (e.g., the presence of spectinomycin, an antibiotic). FIG. 40B depicts an exemplary second step in examining potential resistance mutations. In a second strain of E. coli, we will replace the initial gene of interest with a second gene of interest (GOI2) that confers conditional toxicity (e.g., SacB, which converts sucrose to fructan, a toxic product); we will evolve the drug target to become resistant to the endogenous inhibitor while still retaining its activity. This mutant will prevent expression of the toxic gene. FIG. 40C depicts an exemplary third step in countering resistance mutations. In a third strain of E. coli, we will evolve metabolic pathways that produce molecules that inhibit the mutated drug target. In this way, we predict—and, through our second evolutionary pathway, address—mutations that may lead to resistance to terpenoid-based drugs. Figure 40D depicts an exemplary gene encoding system for detecting Src kinase inhibitors. In short, Src activity enables expression of a toxic gene (GOI2); then, inhibition of Src confers survival. Figure 40E shows an embodiment of the roof of principle for the B2H system described in Figure 40B. The system shown here includes two GOIs: SpecR and SacB. Expression of the GOI confers survival in the presence of spectinomycin; expression of the GOI leads to toxicity in the presence of sucrose. The image depicts the results of growth-coupled assays performed on strains of Escherichia coli in the presence of various concentrations of sucrose. The strain carrying the active form of PTP1B (WT) grows better at high sucrose concentrations than the strain carrying the inactivated form of PTP1B (C215S).

FIG. 41A depicts an exemplary strategy for the evolution of PTP1B inhibitors.

FIG41A depicts an exemplary structural analysis for identifying targets for mutagenesis in the active site of a terpene synthase. It shows the alignment of the class I active sites of ABS (grey, PDB entry 3s9v) and TXS (blue, PDB entry 3p5r) with the positions of the sites highlighted on ABS (red) for targeting site-saturation mutagenesis (SSM). Substrate analogs of TXS (yellow) appear for reference. FIG41B depicts an exemplary strategy for introducing diversity into a library of metabolic pathways: an iterative combination of SSM of key sites on a terpene synthase (as in a), error-prone PCR (ePCR) of the entire terpene synthase gene, SSM of sites on a terpene functionalizing enzyme (e.g., P450), and ePCR of the entire terpene functionalizing enzyme. FIG41C depicts an exemplary quantification of total terpenoids present in DMSO samples of extracts with various TS-containing strains. Briefly, we performed site-saturation mutagenesis at six sites on ADS (similar to the sites shown in FIG. 41A ); we plated the SSM library on agar plates containing different concentrations of spectinomycin; we picked colonies that grew on plates containing high concentrations (800 μg/ml) of spectinomycin and used each colony to inoculate a separate culture; we used a hexane overlay to extract the terpenoids secreted into each culture broth; we dried the hexane extracts on a rotary evaporator and resuspended the solids in DMSO; and we used GC-MS to quantify the total amount of terpenoids present in DMSO. FIG. 41D depicts an exemplary analysis of the inhibitory effects of various extracts on PTP1B. Briefly, the figure shows the initial rate of PTP1B-catalyzed hydrolysis of p-nitrophenyl phosphate (pNPP) in the presence of the terpenoids quantified in FIG. 41C . Two mutants of ADS (G439A and G400L) specifically generated potent PTP1B inhibitors.

FIG42 depicts an exemplary analysis of the link between B2H activation and cell survival. An exemplary strain of E. coli contains both (i) an optimized bacterial two-hybrid (B2H) system ( FIG33E ) and (ii) the terpenoid pathway depicted in FIG36A . Note: pTS includes GGPPS only when ABS or TXS are present; the "Y/F" operon corresponds to a B2H system in which the substrate domain cannot be phosphorylated. Survival at high concentrations of spectinomycin requires activation of the B2H system (i.e., phosphorylation of the substrate domain, a process promoted by inhibition of PTP1B).

FIG43 provides exemplary product profiles of strains of E. coli carrying various terpene synthases. For this figure, strains of E. coli carry (i) an optimized B2H system ( FIG33E ) and (ii) a terpenoid pathway ( FIG36A ). The pathways corresponding to each profile differ only in the composition of the pTS plasmids, which contain TXS (taxene synthase from Taxus brevifolia and geranylgeranyl diphosphate synthase from Taxus canadensis); GHS (γ-humulene synthase from Abies abies); ADS (amorphadiene synthase from Artemisia annua); ABS (abietenyl synthase from Abies abies and geranylgeranyl diphosphate synthase from Taxus canadensis); G400A (G400A mutant of amorphadiene synthase from Artemisia annua) and G439L (G439L mutant of amorphadiene synthase from Artemisia annua). Note that the two mutants of ADS produce products with characteristics different from those of the wild-type enzyme (ADS); our results indicate that the products generated by these two mutants are more inhibitory than those generated by the wild-type enzyme (Figure 41E).

Figures 44A-D provide exemplary structure- and sequence-based evidence supporting the extension of the B2H system to other protein tyrosine phosphatases (PTPs).

图44C证明了PTP1B和PTPN6的催化结构域的示例性序列比对(EMBOSS Needle¹)。图44D提供了PTP1B和TPPN6的催化结构域的示例性序列比较。序列共享34.1％序列同一性和53.5％序列相似性。概括来说，该图的结果指示我们的B2H系统可易于延伸至拥有催化结构域的PTP，该催化结构域(i)结构上类似于PTP1B的催化结构域(这里，我们定义结构相似性为这样的两个结构，当对齐时，具有≤

RMSD的RMSD，框架类似于通过PyMol的对齐功能使用的一个)，和/或(ii)序列类似于PTP1B的催化结构域(这里，我们定义序列相似性为≥34％序列同一性或≥53.5％序列相似性，如由EMBOSS Needle算法定义的)。Figure 44A provides an exemplary structural alignment of PTP1B and PTPN6, two PTPs that are compatible with the B2H system (see Figures 1e and 7 of Update A for evidence of compatibility). We used the alignment function of PyMol to align each structure of PTPN6 with the structure of the catalytic domain of PTP1B either (i) without ligand (3A5J) or (ii) with ligand (2F71). The alignment function performs sequence alignment followed by structural superposition, thus effectively aligning the catalytic domains of the two proteins. Figure 44B provides an exemplary structural comparison of PTP1B and PTPN6; the mean square root deviation (RMSD) of the aligned structures of PTP1B and PTPN6 ranges from 0.75 to 1.15.

Figure 44C demonstrates an exemplary sequence alignment of the catalytic domains of PTP1B and PTPN6 (EMBOSS Needle ¹ ). Figure 44D provides an exemplary sequence comparison of the catalytic domains of PTP1B and TPPN6. The sequences share 34.1% sequence identity and 53.5% sequence similarity. In summary, the results of this figure indicate that our B2H system can be easily extended to PTPs that possess a catalytic domain that is structurally similar to the catalytic domain of PTP1B (here, we define structural similarity as two structures that, when aligned, have ≤

RMSD of the RMSD, framework similar to the one used by the alignment function of PyMol), and/or (ii) sequence similarity to the catalytic domain of PTP1B (here, we defined sequence similarity as ≥34% sequence identity or ≥53.5% sequence similarity as defined by the EMBOSS Needle algorithm).

definition

As used herein, use of the term "operon" can refer to a cluster of genes under the control of a single promoter (as in the classical definition of an operon) and can also refer to a gene encoding system that includes multiple operons (e.g., the bacterial two-hybrid system).

As used herein, a "phosphorylation-regulating enzyme" refers to a protein that regulates phosphorylation.

As used herein, "phosphorylation" refers to a biochemical process involving the addition of phosphate to an organic compound.

As used herein, an "optogenetic actuator" refers to a genetically encodable protein that undergoes a light-induced conformational change.

As used herein, "dynamic range" refers to the ratio of activity in the dark and light states (ie, initial rate in the dark/initial rate in the presence of 455 nm light).

As used herein, "operon" refers to a unit composed of multiple genes that regulate other genes responsible for protein synthesis.

As used herein, "operably linked" refers to one or more genes (i.e., DNA sequences) being appropriately positioned and oriented in a DNA molecule for transcription to be initiated from the same promoter. A DNA sequence operably linked to a promoter means that the expression of the DNA sequence is under the transcriptional initiation regulation of the promoter.

As used herein, "construct" refers to an engineered molecule, such as linked pieces of DNA as a DNA construct; as a RNA construct, one continuous sequence derived from the expression of a DNA construct.

As used herein, "fusion" refers to an engineered constructed expression product, ie, the combination of several linked sequences as one molecule or a single gene encoding a protein-protein fusion originally encoded by two genes.

As used herein, "expression vector" or "expression construct" refers to an operon, plasmid or virus designed for DNA expression of a construct in a host cell, which usually contains a promoter sequence operable in the host cell.

As used herein, "promoter" refers to a DNA region that initiates transcription of a specific DNA sequence. The promoter is located near the transcription start site in the 5' region toward the sense strand. The promoter can be a constitutive promoter, such as the cytomegalovirus (CMV) promoter in mammalian cells, or an inducible promoter, such as a tetracycline-inducible promoter in mammalian cells.

As used herein, "transformation" refers to the delivery of an exogenous nucleic acid sequence or plasmid into a prokaryotic host cell, for example, an expression plasmid (eg, a plasmid expression construct) that is inserted into or taken up by a host cell.

As used herein, "transfection" refers to the insertion of a nucleic acid sequence into a eukaryotic cell.

Transformation and transfection can be transient, so that the nucleic acid sequence or plasmid introduced into the host cell is not permanently incorporated into the cell genome. Stable transformation and transfection refer to the host cell retaining the foreign nucleic acid sequence or plasmid for multiple generations, whether or not the nucleic acid or plasmid is integrated into the host cell's genome.

As used herein, "host" with respect to a cell refers to a cell, such as a bacterium, that is intended to receive a nucleic acid sequence or plasmid or that already carries a nucleic acid sequence or plasmid.

As used herein, "conjugate" refers to the covalent attachment of at least two compounds, for example, a photosensitive element attached to a phosphatase protein.

As used herein, a "bait" with respect to a protein construct is incapable of binding to DNA and/or RNA polymerase.

DETAILED DESCRIPTION

The present invention relates to the field of genetic engineering. In particular, the present invention relates to the construction of operons for producing bioactive agents. For example, operons can be constructed to produce reagents for controlling the functions of biochemical pathway proteins (e.g., protein phosphatases, kinases and/or proteases). Such reagents may include inhibitors and regulators that can be used to study or control phosphatase functions associated with phosphatase pathways or abnormal expression levels. Fusion proteins, such as light-activated protein phosphatases, can be genetically encoded and expressed as light-controlled switch phosphatases. Systems for controlling phosphatase functions in living cells or identifying small molecule inhibitors/activators/regulator molecules of protein phosphatases associated with cell signaling are provided.

The present invention also relates to the assembly of genetically encoded systems (e.g., one or more operons) for detecting and/or constructing bioactive agents. For example, the system can be assembled to accomplish one or more goals, such as (i) to detect and/or synthesize small molecules that affect the activity of regulatory enzymes (e.g., protein phosphatases, kinases, and/or proteases); (ii) to detect and/or evolve regulatory enzymes regulated by light (e.g., light-responsive protein phosphatases, kinases, or proteases), etc. Small molecule modulators can include inhibitors of phosphatases known to be associated with human diseases or involved in causing or maintaining human diseases; phosphatase activators involved in or known to be associated with human diseases (e.g., diabetes, obesity, and cancer); such small molecules can be used as chemical probes in the study of cell signaling; as structural starting points (i.e., guides), etc., for the development of drug compounds for the treatment of human diseases. Photosensitive enzymes can include protein tyrosine phosphatases fused to optogenetic actuators (e.g., if phototropin 1 is a LOV domain). Such fusions can be used as tools for spatiotemporal control of protein tyrosine phosphorylation in living cells.

In addition, microbial operons designed to identify small molecule inhibitors, activators or modulator molecules, light-operated switchable enzymes or biological components (including intracellularly expressed molecules) are provided, including, for example, operons with components for whole-cell microbial screening assay systems. Inhibitor/modulator molecules discovered using the compositions, systems and methods described herein are expected to be used to treat diseases such as diabetes, type II diabetes, obesity, cancer and other disorders associated with protein phosphatases such as Alzheimer's disease.

In one embodiment, the present invention relates to protein tyrosine phosphatase 1B (PTP1B). PTP1B represents a valuable starting point for this research for four reasons: (i) it has been implicated in diabetes ⁵ , obesity ⁶ , cancer ³⁰ , anxiety ³¹ , inflammation ³² , immune response ⁷ and neural specification in embryonic stem cells ³³ , (ii) the mechanism of its subcellular localization is well understood (a short C-terminal anchor links it to the ER; proteolysis of the anchor releases it to the cytosol) ^{29 34} . (iii) it can be easily expressed, purified and analyzed ³⁵ , and (iv) it is a member of a class of structurally similar enzymes (PTPs) that facilitate rapid extension of the architecture to make them photoswitchable. PTP1B represents both an experimentally tractable model system for testing strategies for optical control and an enzyme for which optical modulation is contemplated to allow detailed analysis of a variety of different diseases and physiological processes.

In particular, the exemplary figures: Figures 1, 2, 3, 4, 8, 9, 12 and 13 describe operons that can be used to evolve, improve or optimize optogenetic and imaging technologies (i.e., photosensitive enzymes and genetically encodeable biosensors); Figures 10 and 11 describe strategies for using operons to evolve, improve or optimize photosensitive enzymes; Figures 5, 6, 14, 15, 16, 17, 18, 19, 20, 28, 29, 30 and 31 support (i) the development of operons for detecting and/or evolving small molecules that inhibit known drug targets and (ii) the subsequent characterization of those molecules; Figures 22, 23, 24, 25, 26, 27 and 32 provide examples of the kinetics and biophysical properties of microbially synthesizable molecules that inhibit PTP1B.

I. Disease-related protein tyrosine phosphatases (PTPs) and protein tyrosine kinases (PTKs).

Protein tyrosine phosphatases (PTPs) and protein tyrosine kinases (PTKs) are two classes of enzymes that contribute to aberrant signaling events in various diseases (e.g., diabetes, cancer, atherosclerosis, and Alzheimer's disease, among others) and to understanding disease ^{progression14,36} . In addition, they are involved in regulating memory, fear, appetite, energy expenditure, and metabolism, so the use of such phosphorylation-regulated enzymes can reveal connections between seemingly disparate physiological ^{processes14,22,13} .

Embodiments of light-controlled switch constructs for controlling PTPs and PTKs are described herein. Accordingly, in addition to being used to identify specific alleles (i.e., gene sequences or proteins) of PTPs and/or PTKs associated with specific diseases (such as diabetes, etc., including subtypes of the disease, i.e., early-onset, late-onset, etc., and specific types of cancer) - or other enzymes that they regulate, and for screening and testing molecules (including small molecules) for treating diseases associated with these alleles, the examples of light-controlled switch constructs of PTPs and PTKs developed as described herein should be widely useful for understanding how healthy and diseased cells process chemical signals for interested biomedical researchers.

Although other references describe light control of proteins, including the use of LOV2 conjugates, these references do not mention the use of phosphatases. Fan et al., "Optical Control Of Biological Processes By Light-Switchable Proteins." Wiley Interdiscip Rev Dev Biol. 4(5):545-554. 2015. This reference describes a blue light-oxygen-voltage-sensing (LOV) domain that includes the C-terminal α-helix of LOV2 from oat phototropin, referred to as Jα. Attachment of the LOV domain can cage a protein of interest (POI), while light-induced conformational changes in the LOV domain lead to its uncage. As an example, a peptide kinase inhibitor can be caged by fusion to the C-terminus of LOV2. Exposure to light leads to uncage of inhibitors for light-regulated protein kinase activity in cells. WO2011133493. "Allosteric regulation of kinase activity." published on October 27, 2011. This reference describes fusion proteins comprising a kinase (including as examples tyrosine kinase (Src), serine/threonine kinase (p38)) and a ligand binding domain, such as a light-regulated LOV domain (wherein the illumination of ligand binding is taken into account), which is inserted into the N-terminus and/or C-terminus or near the catalytic domain to produce allosteric regulation using a light-dependent kinase. In addition, the LOV domain includes a LOV2 domain and/or a Ja domain from oat phototropin I. WO2012111772 (A1) in Japanese, abstract in English. This reference abstract describes a polypeptide for optical control of calcium signaling comprising an amino acid sequence comprising the LOV2 domain of SEQ ID NO: 1 or an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 1. The construct has a LOV2 domain followed by a LOV2-Jα photoswitch at the N-terminus of the construct. US8859232. "Genetically encoded photomanipulation of protein and peptide activity.", published on October 14, 2014. This reference describes fusion proteins including protein photoswitches, and methods of light-manipulating the activity of fusion proteins to study protein function and analyze subcellular activity, as well as diagnostic and therapeutic methods. More particularly, the fusion protein includes a protein of interest fused to a protein photoswitch including a light, oxygen, or voltage (LOV2) domain of oat (oat) phototropin 1, wherein illumination of the fusion protein activates the protein of interest or inactivates the protein of interest. The protein of interest is a functional domain of a human protein. As an example, the LOV2-Jα sequence of phototropin 1 (404-547) is fused to the N-terminus of Racl so that the LOV domain in its closed conformation will reversibly block binding of the effector to Racl.

A. Protein tyrosine phosphatase (PTP).

Protein tyrosine phosphatases (PTPs) are a class of regulatory enzymes that exhibit abnormal activity in their various diseases. Therefore, the detailed drawing of the allosteric communication of these enzymes can reveal the structural basis of physiologically relevant-and possibly therapeutically useful-perturbations (i.e., mutations, post-translational modifications, or binding events) that affect their catalytic state. This study combines the detailed biophysical study of protein tyrosine phosphatases IB (PTP IB) with large-scale bioinformatics analysis to examine the allosteric communication in PTPs. The results of X-ray crystallography, molecular dynamics simulations, and sequence-based statistical analysis indicate that PTP IB has a widely distributed allosteric network that is evolutionarily conservative throughout the PTP family, and the findings from dynamics studies show that the network is functionally complete in sequence-diverse PTPs. The allosteric network solved in this study reveals a new site for targeting PTP allosteric inhibitors, and helps to explain the functional impact of various disease-related mutations.

In one embodiment, tyrosine phosphatases and photosensitive proteins as described herein can be attached to drugs for medical treatment. In contrast to EP2116263, "Reversibly light-switchable drug-conjugates." published on November 11, 2009, does not mention tyrosine phosphatases, and describes a light-switchable conjugate of the protein phosphatase calcineurin attached to a photoisomerization group B and also attached to a drug for medical treatment (none of these groups is genetically encodable). As an example in EP2116263, tumor growth is inhibited by inhibiting the protein phosphatase calcineurin. The photoisomerization group B for activity induced by near UV (e.g. 370nm) or near IR (e.g. 740nm) does not include the N-terminal alpha helix of the photoresponsive plant protein phototropin 1LOV2.

Receptor PTPs are contemplated for conjugation to light-sensitive proteins, as described herein. In contrast, Karunarathne et al., "Subcellular optogenetics-controlling signaling and single-cell behavior." J Cell Sci. 128(1): 15-25, 2015, describe light-sensitive domains, such as bacterial light-oxygen-voltage-sensing (LOV and LOV2) domains, including a C-terminal helical Jα region, which is linked to receptor tyrosine kinases (RTKs), without specific examples, and without mentioning tyrosine phosphatases or the plant phototropin 1 LOV2 N-terminal α helix. Photoactivation of inositol 5-phosphatases is shown, but inositol 5-phosphatases are not protein phosphatases.

B. Enzymatic phosphorylation of tyrosine residues.

Enzymatic phosphorylation of tyrosine residues plays a role in cellular function and is aberrantly regulated in a variety of diseases, such as diabetes, cancer, autoimmune disorders, and Noonan syndrome. It is controlled by the concerted action of two classes of structurally flexible—and dynamically regulatable—enzymes: protein tyrosine kinases (PTKs), which catalyze the ATP-dependent phosphorylation of tyrosine residues, and protein tyrosine phosphatases (PTPs), which catalyze the hydrolytic dephosphorylation of phosphotyrosine (5, 6). A detailed understanding of the mechanisms by which these enzymes respond to activity-regulated structural perturbations (i.e., mutations, post-translational modifications, or binding events) will elucidate their contributions to various diseases and facilitate the design of new PTK- or PTP-targeted therapies. Over the past several decades, numerous biophysical studies have dissected the catalytic mechanisms and regulatory functions of PTKs, which are common targets for pharmaceutical agents (7, 8). (9) In contrast, detailed analysis of PTPs has lagged behind. (10) These enzymes represent an untapped source of biomedical insights and therapeutic potential (no PTP inhibitors have cleared clinical trials); therefore, they are the focus of this research.

PTP uses two loops to dephosphorylate tyrosine residues. The eight-residue P-loop binds a phosphate moiety via a positively charged arginine, which enables nucleophilic attack by a nearby cysteine, and the ten-residue WPD loop contains a common acid catalyst—aspartic acid—that protonates the tyrosine leaving group and hydrolyzes the phosphatase intermediate (11-13). During catalysis, the P-loop remains fixed, while the WPD loop moves ~10 Å between open and closed states; nuclear magnetic resonance (NMR) analysis shows that this movement controls the rate of catalysis (14).

Recent analysis of protein tyrosine phosphatase IB (PTP IB), a drug target for the treatment of diabetes, obesity, and breast cancer, indicates that the motion of its WPD loop is regulated by an allosteric network extending to its C-terminus (Fig. 1B) (15, 16). This network is susceptible to modulation by both (i) inhibitors that replace its C-terminal a7 helix (17, 18) and (ii) mutations that disrupt communication between the (α)7 helix and the WPD loop (15); the specific collection of residues that enable allosteric communication in PTP1B and other PTPs remains incompletely resolved.

Protein tyrosine phosphatase 1B (PTP1B). PTP1B represents a valuable tool for identifying potential therapeutics for at least four reasons: (i) it has been implicated in ^diabetes5 , ^obesity6 , ^cancer30 , ^anxiety31 , ^{inflammation32} , immune ^responses7, and neural specification of embryonic stem ^cells33 , (ii) the mechanisms underlying its subcellular localization are well understood (a short C-terminal anchor links it to the ER; proteolysis of this anchor releases it into the cytosol) ^2934. (iii) it can be readily expressed, purified, and ^analyzed35 , and (iv) it is a member of a class of structurally similar enzymes (PTPs) that promote rapid extension of their framework, making them photoswitchable. Therefore, PTP1B represents an experimentally tractable model system for testing strategies for optical control, and an enzyme whose optical modulation allows detailed analysis of a variety of different diseases and physiological processes.

Spatial regulation and intracellular signaling. By way of example, PTP1B demonstrates the value of photoswitchable enzymes for studying spatial regulation in intracellular signaling. Receptor tyrosine kinases are hypothesized to be inactivated by: (i) contacts between endosomes and the ^ER37,38 , (ii) contacts between the plasma membrane and extended regions of the ^ER39 , and (iii) direct protein-protein interactions through their partial proteolysis and release into the ^cytosol34 . The role of different mechanisms (or locations) of PTP1B-substrate interactions in determining the outcome of those interactions is poorly understood. Evidence suggesting a relationship between the location of PTP1B and its role in signaling has emerged from studies of tumorigenesis. Inhibition of PTP1B suppresses tumor growth and metastasis in breast ^cancer30,40 , lung ^cancer3,41 , colorectal cancer9 ^, and prostate ^cancer42,43 , while its upregulation has similar effects in ^lymphoma3,44 . Recent evidence suggests that the former effect may be caused by inhibition of cytoplasmic ^PTP1B45 ; the cause of the latter is unclear. Currently, there are no tools to investigate the differential effects of spatially distinct PTP1B subpopulations on tumor-associated signaling events within the same cell. Photoswitchable variants of PTP1B represent such a tool.

Network biology. Signaling networks typically represent nodes (proteins) connected by lines (interactions) ⁴⁶ . Such diagrams capture the connectivity of biochemical relay systems but obscure spatial context—the ability of a single interaction to occur at multiple locations and potentially stimulate multiple signaling effects. This study develops a set of tools that will enable detailed investigation of the role of spatial context in directing the propagation of signals through biochemical networks; such examinations are useful for understanding the role of PTP1B in cellular signaling (and processes relevant to tumorigenesis) and are generally relevant to the study of any enzyme present in spatially distinct subpopulations within cells.

II. Optogenetic Actuators.

Optogenetic actuators (genetically encodable proteins that undergo light-induced conformational changes) provide a convenient way to orchestrate biochemical events under optical control. Alone or when fused to other proteins, they enable optical manipulation of biomolecular transport, binding, and catalysis in living cells with millisecond and submicron resolution. Our approach addresses two major flaws in existing techniques: observational interference and illumination half the story. Existing strategies for controlling enzyme activity with light interfere with the natural patterns of protein production, localization, and interaction (usually by design), so it is difficult to directly interrogate and/or control those patterns, which determine how biochemical signals are processed. There are several methods for controlling protein kinases with light, but there are no similar methods for controlling protein phosphatases. Because signal transduction networks are regulated by the synergistic action of two classes of enzymes, comprehensive control and/or detailed analysis of those networks requires two control methods.

Embodiments described herein include: (i) methods for controlling protein activity with light without destroying its wild-type activity and (ii) demonstration of this method on a particularly important protein: protein tyrosine phosphatase 1B (PTP1B), a regulator of cell signaling and a validated drug target for the treatment of diabetes, obesity, and cancer. There are no known photoswitchable protein tyrosine phosphatases. The PTP1B-LOV2 construct reported in this document is the first. (ii) The N-terminal alpha helix of LOV2 has been ignored in most studies (even reviews of photoswitches) and has not been used as an exclusive attachment point for light modulation of the enzyme.

We developed a photoswitchable version of PTP1B by fusing the enzyme's C-terminal allosteric domain to the N-terminal alpha helix of a protein photoswitch (i.e., the LOV2 domain of phototropin 1 from oats). We present evidence that this general architecture—unique in that the LOV2 position is distal (minimally disruptive) to the active site of PTP1B—can be extended to other PTPs, and possibly PTKs. For example, we used statistical coupling analysis to show that the allosteric network exploited in our PTP1B design is conserved across the entire PTP family.

Alone or when fused to other proteins, optogenetic actuators enable optical manipulation of biomolecular transport, binding, and catalysis with millisecond and submicrometer ^{resolution15,16} . At least three drawbacks limit their use for detailed studies of signaling networks: Observational perturbations. Existing strategies for controlling enzyme activity with light perturb the natural patterns of protein production, localization, and ^{interaction16,17} (usually by design), so direct interrogation of those patterns—which determine how biochemical signals are ^processed10 —is difficult. Illumination halves the odds. There are several methods for controlling protein kinases ^{with light18,19} , but no similar methods for controlling protein phosphatases. Because signaling networks are regulated by the concerted action of two classes of enzymes, detailed analysis of those networks requires two methods of control. A limited palette of actuators. Optogenetic actuators that enable subcellular control of enzyme activity require the use of blue or green ^light15 . These wavelengths exhibit significant phototoxicity ²⁰ , suffer from short biological penetration depths ²¹ , and, due to their spectral similarity, limit actuation to single signaling events rather than simultaneous multiple events.

A. Photoswitch constructs: advantages over other exemplary technologies.

As described herein, the light-operated switch describes a protein-protein framework (e.g., PTP1B-LOV2 fusion) that is optically active in its monomeric form. Reference, WO2013016693. "Near-infrared light-activated proteins." Publication date January 31, 2013, relies on homodimerization. In contrast, unlike in WO2013016693, the optical control described herein exceeds a larger range of proteins, including both those that require homodimerization and those that do not require homodimerization. In addition, the reference describes types of photosensitive modules, including blue light-sensitive flavoproteins found in plants; blue light photoreceptors using flavin adenine dinucleotide (BLUF); light, oxygen or voltage sensing (LOV) types, including plant and bacterial photoreceptors; and plant/microorganism phytochromes that are sensitive to light, i.e., light-induced helical rotations in the red-to-NIR region. The examples more particularly describe fusion proteins based on the light activation of bacterial phytochromes (Bph), using a light-responsive alpha helix from Rhodobacter sphaeroides (BphG) fused to proteins such as protein phosphatases, protein kinases, membrane receptors, etc. E. coli were modified so as to display the level of photoactivity of these expressed fusion proteins, i.e. in the presence or absence of red-to-NIR light. Although blue color changes in E. coli expressing fusion proteins were described in response to light, these blue bacteria were the result of far-red/NIR-light using photoactivated fusion proteins, which in turn activated lacZ expression in the presence of Xgal, rather than photoresponse to exposure to blue light. However, there is no specific mention of tyrosine phosphatases or plant phototropins 1 LOV2 N-terminal alpha helix. In fact, reviews on optogenetics tend to describe LOV2 as having one terminal helix: the C-terminal Jα helix. While there are studies/patents indicating that insertion of simple LOV2 domains enables light control, they rely on the assumption that the Jα helix is unwinding to produce the control effect, not the Aα helix as described herein.

B. A "cage-free" approach to control protein tyrosine phosphatases and protein tyrosine kinases using light.

Current strategies for using light to control enzyme activity (as opposed to its concentration or location) rely on cage-based systems: light-responsive proteins, when fused to an enzyme of interest, control access to its active ^site16,47 . Unfortunately, this architecture can alter the enzyme's affinity for binding partners and change its sensitivity to activity-regulating modifications (e.g., phosphorylation) ^16,18 . These effects complicate the study of signaling using optogenetics. This study will develop a "cage-free," allosteric-based optical control approach that minimizes interference between the enzyme and its substrates (and other binding partners). The approach will help preserve native patterns of protein localization, interactions, and post-translational modifications, thereby facilitating studies of the effects of those patterns on intracellular signaling.

2. Genetically encoded photoswitchable phosphatases. No genes encode photoswitchable phosphatases; the chimeras developed in this proposal would be the first. Photoswitchable variants of PTP1B would enable detailed studies of a variety of PTP1B-regulated processes of interest (e.g., insulin, endocannabinoid, and epidermal growth factor ^{signaling49,51} , and cell adhesion and ^migration52 ). In general, photoswitchable phosphatases would provide a useful class of tools for studying cell biology (particularly in concert with photoswitchable kinases, which would enable complementary experiments).

Hypothesis: The catalytic domains of PTPs and PTKs possess a C-terminal α-helix that is distal to their active site yet capable of modulating their catalytic activity (for at least a subset of enzymes—the universality of this function is unknown) ^{23, 24.} We hypothesized that fusion of this helix with the N-terminal α-helix of the light-oxygen voltage 2 (LOV2) domain from oat phototropin 1—a photosensitive domain with a terminal helix that unwinds in response to blue light ^{25, 26} —would produce an enzyme-LOV2 chimera that exhibits light-dependent catalytic activity yet retains its native substrate specificity and binding affinity.

Experimental approach: We attach the C-terminal a-helix of PTP1B to the N-terminal a-helix of LOV2 at a homologous junction, and we will assess the effect of photoactivation on the catalytic activity of the resulting chimera. This effort will involve the use of (i) kinetic assays and binding studies to characterize the substrate specificity and binding affinity of the photoswitch constructs and (ii) crystallographic and spectroscopic analyses to examine the structural basis of photocontrol. Informed by these studies, we extend our approach to striatal-enriched protein tyrosine phosphatases (STEP) and protein tyrosine kinase 6 (PTK6), enzymes implicated in Alzheimer's disease and triple-negative breast cancer, respectively.

We will combine precision biophysical studies, synthetic biology, and fluorescence microscopy to (i) develop protein architectures that enable optical control of protein tyrosine phosphatases (PTPs) and protein tyrosine kinases (PTKs) without perturbing their wild-type activity or binding specificity, (ii) evolve PTPs and PTKs that are regulated by red light, and (iii) develop imaging methods to study spatially localized signaling events in living cells.

We will begin our studies with PTP1B, an enzyme that is a validated drug target for the treatment of diabetes, obesity, and breast cancer and for which optogenetic tools will be particularly useful to address gaps in current knowledge (e.g., the role of spatially distinct PTP1B subpopulations in promoting or inhibiting tumor growth ²² ). Using this as a model, we will establish the generalizability of our approaches by extending them to other PTPs and PTKs.

C. Photoswitchable variants of PTP1B.

Our first goal sought to use LOV2, a protein with a terminal helix that unwinds in response to blue light, to control the activity of PTP1B, an enzyme whose unwinding of the C-terminal a-helix destroys activity by twisting its catalytically essential WPD loop (Fig. 1AB, Fig. 6). To assess the feasibility of this goal, we constructed five PTP1B-LOV2 chimeras (linked at homologous intersections): three chimeras showed light-dependent catalytic activity toward 4-methylumbelliferyl phosphate (4M) (Fig. 1G). Subsequent mutational analysis of one chimera indicated that mutations in the a-helix that connects PTP1B to LOV2 improved catalytic activity and dynamic range (DR, ratio of dark/light activity; Fig. 1G). Our ability to construct—and begin to optimize—a photoswitchable PTP1B-LOV2 chimera by screening a small number of constructs suggested that rational design would allow us to construct chimeras sufficient for intracellular signaling studies. We note that most of our photoswitchable chimeras have a DR of 2.2; previous imaging studies have shown that a DR of 3–10 is sufficient to control intracellular ^{signaling218,19} .

More specifically, Figure 1C demonstrates some differences relative to other types of optical control. The y-axis of the top graph indicates the activity of each construct in the dark (i.e., the initial rate of p-nitrophenyl phosphate hydrolysis catalyzed by PTP1B); the y-axis of the bottom graph indicates the ratio of activity in the dark and light states (i.e., initial rate in the dark/initial rate in the presence of 455 nm light), i.e., the dynamic range.

The black bars show the activity and dynamic range of a set of eight initial constructs with different crossing points (see bottom of Figure 1B). Some of these constructs are photoswitchable and some are not. Version 7 showed the maximum photoswitchability - the dynamic range is about 1.8.

More specifically, the colors are associated with different types of constructs. Black: different junctions (for junctions, see Figure 1B); Gray: different partitions of the linker (see, Linker section below); Light blue: Jα helix—this is at the C-terminus of the LOV2 domain; Dark blue: A'α helix—this is at the N-terminus of the LOV2 domain, and therefore, in the region that connects it to PTP1B; Yellow: α7 helix of PTP1B—this is at the C-terminus of PTP1B, and therefore, in the region that connects PTP1B to LOV2; Orange: Combination: combination of sites from previous colors, see below for additional information.

These results are surprising, in part because a recent review of optogenetics showed that light-controlled activity requires the Jα helix of LOV2 (where Jα is a C-terminal helix that resides in a folded state against the LOV domain core) to be attached to the protein of interest, see Repina, N.A., Rosenbloom, A., Mukherjee, A., Schaffer, D.V. & Kane, R.S. At Light Speed: Advances in Optogenetic Systems for Regulating Cell Signaling and Behavior. Annu. Rev. Chem. Biomol. Eng. 8, 13–39 (2017). Photoactivation with blue light converts the non-covalent interaction between the LOV core and its bound flavin chromophore FMN into a covalent interaction through a conserved cysteine residue. The accompanying light-induced conformational change moves the Jα helix away from the protein core, resulting in the uncaging of the fused effector domain (e.g., the kinase domain of phot1). The Jα helix returns to its dark state cage-like conformation within minutes due to spontaneous decay of the protein-cofactor bond.

Several limitations of the native AsLOV2 domain have prompted efforts to engineer improved variants. First, spontaneous undocking of the Jα helix when fused to an exogenous protein domain can result in relatively high dark state activity, leading to a low dynamic range after AsLOV2 uncaging (26). For example, the light-induced DNA binding system LovTAP has only a five-fold change in DNA affinity between the dark and illuminated states (27). To address this issue, Strickland et al. (26) used rational design to introduce four mutations into AsLOV2, which stabilized the docking of Jα to the LOV core. This increased the dynamic range of LovTAP from 5-fold to 70-fold, and the method can be applied to other LOV domain optogenetic systems to reduce dark state activity. AsLOV2 fusions are also particularly sensitive to linker length and the size and structure of the attached domain (28, 29), so each new fusion protein switch needs to be optimized to achieve low dark state and high light state activity in mammalian cells.

In contrast to Jα helix-protein chimeras, as shown herein, the A'α helix rather than the Jα helix is attached to the protein of interest to form a photoswitchable construct, such as PTPB1.

Exemplary linkers.

The grey bars in Figure 1C show the activity and dynamic range of mutants of version 7, in which the linker has been repartitioned. In other words, version 7 has the following linker region: LSHEDLATTL, where the underlined region "LSHED" corresponds to the C-terminus of PTP1B, and the region "LATTL" corresponds to the N-terminus of LOV2. Version 7.1 has the sequence LSHEDATTL; version 7.2 has the sequence LSHEDTTL, etc. Here, we found that version 7.1 has the same dynamic range as version 7, but with higher activity. Therefore, we used version 7.1 for further optimization.

Exemplary mutants.

The light blue bars show the activity and dynamic range of mutants of version 7.1 that contain helix-stabilizing mutations in the Jα helix. Curiously, these improved the activity of 7.1 but not its dynamic range.

The dark blue bars show the activity and dynamic range of mutants of version 7.1, in which the A' α helix contains helix-stabilizing mutations. One of these mutations (T406A) improved the dynamic range; we used this version for further studies.

The yellow bars show the activity and dynamic range of mutants of version 7.1, in which α7 of PTP1B has a helix-stabilizing mutation; the orange bars show the activity and dynamic range of mutants of version 7.1, in which multiple mutations are combined. None of the constructs associated with the yellow and orange bars showed improved properties of 7.1(T406A).

Minimally disruptive approach. Two kinetic studies indicate that the framework we used for light control does not interfere with the native substrate specificity or binding behavior of PTP1B: (i) Activity analysis of chimeric E3 (from Figure 1D) on p-nitrophenyl phosphate (pN) indicated that light affected _kcat , but not _Km (Figures 2K and L). (ii) Activity analysis on three substrates of different sizes (4M, pN, and peptide) showed that the DRs were the same for all three (Figures 2L-K). The results of both studies are consistent with our hypothesized mechanism of light control: LOV2-induced unwinding of the C-terminal α-helix of PTP1B disrupts the movement of its catalytically essential WPD loop, which controls the rate of catalysis but has little effect on substrate binding affinity.

Biophysical studies. The titer (-100 mg/L) of the photoswitch chimera expression was sufficient for detailed biophysical analysis. We performed a preliminary set of these analyses on chimera E3. (i) We used circular dichroism (CD) to examine the effects of photoactivation on its secondary structure; spectral measurements indicated that photoactivation reduced the a-helix content (222 nm; FIG. 2B ). (ii) We used the amplitude at 222 nm to measure the post-activation recovery time of the a-helix content: T _r -30s ( FIG. 2E ). This value is similar to the recovery time of the previously developed LOV2-based photoswitch construct, (iii) we used tryptophan fluorescence to measure the post-activation recovery time of the tryptophan residue: T _r -50s ( FIG. 2F ). Tryptophan fluorescence is a rough measure of the PTP1B conformation (which has seven tryptophan residues compared to the one in LOV2). Therefore, this slower recovery time indicates that PTP1B takes longer to refold than LOV2. (iv) We identified a set of crystallization conditions (those previously used to crystallize PTP1 B _W t) to grow crystals of E3 (Fig. 2F). (V) We collected a two-dimensional ¹ H- ¹⁵ N HSQC spectrum of PTP1 B _W t and assigned a non-proline peak at -65%. These recent NMR experiments must include the PTP1B-LOV2 chimera; but our ease in performing them (in one attempt) suggests that similar analysis of the chimera is straightforward. The experimental tractability of the PTP1B-LOV2 chimera will enable comprehensive biophysical analysis of variants with different photophysical properties.

Example 1. Development of a "cage-free" approach to control protein tyrosine phosphatases and kinases with light. This section develops methods to place enzymes under optical control without disrupting their natural interactions. We will demonstrate this approach with PTP1B and then extend it to STEP and PTK6. When we have a PTP1B-LOV2 chimera that exhibits a three- to ten-fold change in activity between the light and dark states, and when we identify structure-based design rules that facilitate fine-tuning of the photophysical properties of light-operated switch variants of PTP1B, STEP, and PTK6, we will know that we are successful.

D. Development of photoswitchable variants of PTP1B.

The work in this section assumes—and attempts to confirm with crystallographic, kinetic, and binding studies—that optogenetic actuation systems located far from the active site are unlikely to disrupt the wild-type behavior of actuation systems located nearby. Kinetic studies of preliminary PTP1B-LOV2 chimeras (i.e., chimeras in which the C-terminal helix of PTP1B is linked to the N-terminal helix of the LOV2 domain of phototropin 1 from oats) support this hypothesis: light inhibits its activity by affecting k _cat rather than K _m , and they display wild-type kinetics on 4-methylumbelliferyl phosphate (4M), a model substrate (Figure 1G and Figure 2K). Light modulation of k _cat rather than K _m suggests that LOV2 utilizes an allosteric network to distort the WPD loop (Figure 6).

Our initial constructs, which represent the first reported example of a photoswitchable protein phosphatase, will facilitate systematic studies of the functional advantages of different chimera architectures. We are particularly interested in understanding how (i) the length of the linker connecting PTP1B and LOV2 and (ii) the stability of the terminal helix of LOV2 affect catalytic activity and dynamic range. We will investigate these relationships by combining spectroscopic analysis with kinetic studies. Spectroscopic analysis will show how different PTP1B-LOV2 chimeras rearrange under illumination (e.g., we will use CD and fluorescence spectroscopy to measure photomodulation of a-helical content and tryptophan fluorescence), and kinetic studies will reveal the effects of those rearrangements on catalytic activity and dynamic range.

The results of our biophysical analyses will help optimize our chimeras for in vitro cellular studies. We will target chimeras (hereafter referred to as PTP1B _PS ) with the following properties: a dynamic range (DR) of 3-10 s, a recovery time of _Tr ∼15-60 s relative to wild-type activity (in its activated state). Previous optogenetic studies have shown that these properties enable optical control of cellular ^{signaling2,18,19} . We note that biophysical studies of PTP1B indicate that removal of its C-terminal α-helix reduces its activity by a factor of four ⁵⁷ ; accordingly, we believe that LOV2 may modulate the activity of PTP1B by at least fourfold (of course, LOV2 may trigger structural distortions that are more pronounced than those of simple truncations).

E. Characterization of PTP1B-substrate and PTP1B-protein interactions.

We will evaluate the effect of LOV2 on the substrate specificity of PTP1B by using kinetic analysis. In particular, we will compare the activity of PTP1BWT and PTP1BPS on the following three substrates: (i) p-nitrophenyl phosphate, a general substrate of tyrosine phosphatases, (ii) ETGTEepYMKMDLG, a PTP substrate closely related to PTP1B, and (iii) RRLIEDAEpYAARG, a substrate specific for PTP1B. Comparison of the _kcat and _Km values of these substrates (Figure 2K shows an example kinetic study) will reveal differences in the catalytic activity and specificity of _PTP1BWT and _PTP1BPS . These studies will also allow us to evaluate the substrate dependence of the light-controlled switching ability (i.e., DR). It is usually assumed that light modulation is independent of the substrate. However, this assumption has no biochemical basis (especially in cage-based systems, where substrates can bind with different affinities and therefore have different abilities to compete with cage proteins). We will test it.

We will assess the ability of PTP1B _PS to participate in the same protein-protein interactions as PTP1B _WT by measuring the affinity of both enzymes for two natural binding partners of PTP1B _WT : LM04 and Stat3. Binding isotherms based on tryptophan fluorescence changes of PTP1B will facilitate this study (Figure 7).

The biochemical comparison of our PTP1B _WT and PTP1B _PS may seem tedious, but we feel that this analysis is necessary to establish the relevance of future optogenetic observations to wild-type processes.

Biostructural Characterization. We will investigate the structural basis of photocontrol in PTP1B _PS by using X-ray crystallography and NMR spectroscopy. X-ray crystal structures will show how LOV2 affects the structure of PTP1B (and vice versa); NMR spectroscopy will show how LOV2 modulates catalytic activity. For crystallographic studies, we will crystallize PTP1B _PS in its dark state (we will use the C450S mutation, which prevents the formation of cysteine ^adducts2,26 ) by screening crystallization conditions previously used for LOV2, PTP1B, and LOV2-protein chimeras (all of which have crystal ^{structures2,35,58} ); preliminary results show that those used to grow PTP1B _WT crystals also produce crystals of the PTP1B-LOV2 chimera (Figure 2J). For NMR studies, we will use heteronuclear single quantum coherence (HSQC) spectroscopy and transverse relaxation optimization spectroscopy (TROSY) to monitor changes in the conformation and backbone dynamics of PTP1B _PS before and after irradiation. (We note that backbone ¹ H, ¹³ C, and ¹⁵ N chemical shifts have been assigned to PTP1B and LOV2 ^{59 , 60} ).

G. Exemplary imaging methods for studying subcellular signaling events in living cells.

This section uses the PTP1B _PS (PTP1B-LOV2 chimera) to develop methods to probe and study subcellular signaling events using confocal microscopy. We will know that this goal is successful when we can inactivate subcellular regions, monitor the effects of this inactivation with a FRET-based sensor, and separate the contributions of different subpopulations of PTP1B (e.g., ER-bound, cytoplasmic) to sensor phosphorylation.

Hypothesis. The subcellular localization of PTPs and PTKs is controlled by domains close to their catalytic ^cores23,24 . We hypothesized that attachment of these domains to photoswitchable chimeras would confer their wild-type localization pattern and enable the study of the contribution of spatially distinct subpopulations of PTPs and PTKs to cellular signaling using confocal microscopy. Experimental Approach: Within cells, PTP1B exists in two spatially distinct subpopulations: attached to the cytoplasmic surface of the endoplasmic reticulum, and free in the cytosol—the result of proteolysis of its short (-80 residues) C-terminal ER anchor. We will (i) attach the ER anchor of wild-type PTP1B (PTP1 B _WT ) to our PTP1B-LOV2 chimeras, (ii) compare the subcellular localization of the resulting chimeras with that of PTP1B _WT chimeras, (iii) use confocal microscopy—in conjunction with a FRET-based sensor for phosphatase activity—to control and monitor PTP1B activity within cells, and (iv) develop a reaction-diffusion model to assess the contribution of spatially distinct subpopulations of PTP1B to temporal and spatial changes in sensor phosphorylation. This work will generate a general approach for studying spatially localized signaling events in living cells.

Localization of PTP1B _PS .

To examine the localization of PTP1B _PS in living cells, we will express three variants in COS-7 cells (i) PTP1B _PS _ _C 45os, (ii) PTP1B _PS _ c45os attached to a short fragment (-20 amino acids ²⁹ ) of the C-terminal ER anchor of PTP1B _WT containing only the transmembrane domain (but not the proteolytic site), and (iii) PTP1 B _PS -c45os attached to the entire C-terminal ER anchor of PTP1B _WT (-80 amino acids ²⁹ ). We hypothesize that these constructs will have (i) a cytoplasmic localization pattern, (ii) an ER-bound localization pattern, and (iii) both a cytoplasmic and ER-bound (i.e., wild-type) localization pattern, respectively. Using confocal microscopy, we will test this hypothesis by using the fluorescence of LOV2 to localize each chimera ⁷⁰ . (In these studies, we will localize the ER using fluorescently labeled SEC61B, an ER-associated transport complex. ⁷¹ Locking the C450S mutation of LOV2 in its fluorescent state will prevent photoactivation during imaging).

COS-7 cells, fibroblast-like cells derived from African green monkey kidney tissue, are particularly compatible with the above analysis for three reasons: (i) they are large and flat, thus facilitating imaging of spatially separated subcellular ^regions72 , (ii) they are ^compatible with commercially available transfection reagents73, and (iii) methods for inducing ^{endocytosis71} and calpain ^expression74 (both processes that affect subcellular activity and localization of PTP1B) are well developed for these cells.

Control of PTP1B _PS in living cells. We will examine the activity of PTP1B in subcellular regions by confocal microscopy paired with a FRET-based sensor for protein phosphorylation (developed by the Umezawa group ⁵⁴ ; FIG. 12 ). The sensor will consist of a kinase substrate domain, a short flexible linker, and a phosphorylation recognition domain—all sandwiched between two fluorescent proteins (Clover, green fluorescent protein and mRuby2, red fluorescent protein). Phosphorylation of the substrate domain will cause it to bind to the recognition domain, thereby modulating (i.e., enhancing or reducing) the FRET between the two fluorescent proteins. Our preliminary sensor using substrates and SH2 domains compatible with PTP1B and src ^{23 , 55} exhibited a 20% change in FRET in response to phosphorylation. We will attempt to further optimize our sensor by screening different substrate domains, SH2 domains, and linker lengths. Ouyang et al. established a FRET sensor for Src kinase activity that exhibited a -120% FRET change when phosphorylated ⁵⁵ . We will use the architecture of this sensor—or possibly the sensor itself—to report on our design.

In our imaging experiments, we will use a 455nm laser to inactivate PTP1B within subcellular regions (1-10 μm circles) and fluorescence lifetime imaging microscopy (FLIM) to monitor changes in sensor phosphorylation caused by inactivation (Figure 13). For these experiments, we will use siRNA to deplete PTP1 B _WT and SEC61B to label the ER. The output will be a series of images where the intensity of the pixel is proportional to the fluorescence lifetime of Clover (and, therefore, the extent of sensor phosphorylation).

With this study, we are particularly interested in studying the relationship between (i) the location of activation/inactivation of PTPIBps, (ii) the size of the activation/inactivation region, and (iii) the location and time of changes in the phosphorylation state of the sensor. We will use the reaction-diffusion model to study these relationships. Equation 1 provides a simple example of the governing equation:

Brave with phosphorylated sensor ( _Sp ). Here, _Ds is the diffusion coefficient of the sensor; Ks is the concentration of tyrosine kinase bound to the unphosphorylated sensor; _Psp is the concentration of PTP1B bound to the phosphorylated sensor; P and _Sp are the concentrations of free PTP1B and free phosphorylated sensor, respectively; k^ _at and k^ _at are the catalytic constants of the tyrosine kinase and PTP1B, respectively; and k% _n is the kinetic constant for sensor-PTP1B association. Kinases and phosphatases are assumed to bind only weakly to their products (an assumption that can be easily re-examined later). We can also supplement this model with tools such as BioNetGen, a web-based platform for generating biochemical interactions from user-specified rules for the mechanism and location of biomolecular ^{interactions75} ; such tools that can adapt to cellular heterogeneity (e.g., organelles and other compartments) will help to support and extend our kinetic model.

We hypothesized that in the presence of cytosolic PTP1B _PS , a version of our kinetic model in which the phosphatase diffuses freely will more accurately capture the phosphorylation state of the sensor (at a specified time and location from the radiation region). In contrast, in the presence of ER-bound PTP1B _PS , a version of the model in which the phosphatase cannot diffuse freely will more accurately capture the behavior of the sensor. Regression of either model against imaging data will enable assessment of the extent to which cytosolic and ER-bound PTP1B contribute to changes in sensor phosphorylation over time and space.

Image Analysis. The ER exists as a vesicular network throughout the cell; inactivating subcellular regions that are either entirely ER or entirely cytosol is difficult. To be able to analyze spatially distinct subpopulations of PTP1B, we must therefore assess the amount of ER in different regions that are irradiated. Differences in ER heterogeneity (-20-100 pm) and length scales of irradiation (-1-10 pm) will allow for such assessments. We will work with two metrics: (i) total fluorescence of labeled ER and (ii) anisotropy of labeled ER. By facilitating assessment of populations of cytosolic and ER-bound PTP1B in the illuminated region, these two metrics will help us assess the contribution of those populations to changes in sensor phosphorylation.

Spatial regulation and intracellular signaling.

By way of example, PTP1B illustrates the value of photoswitchable enzymes for studying the spatial regulation of intracellular signaling. Receptor tyrosine kinases are hypothesized to be inactivated by: (i) contacts between endosomes and the ^ER37,38 , (ii) contacts between the plasma membrane and extended regions of the ^ER39 , and (iii) enabling direct protein-protein interactions through their partial proteolysis and release into the ^cytosol34 . The role of different mechanisms (or locations) of PTP1B-substrate interactions in determining the outcome of those interactions is poorly understood. Evidence suggesting a relationship between the location of PTP1B and its role in signaling has emerged from studies of tumorigenesis. Inhibition of PTP1B suppresses tumor growth and metastasis in breast ^cancer30,40 , lung ^cancer3,41 , colorectal ^cancer9 , and prostate ^cancer42,43 , while its upregulation has similar effects in ^lymphoma3,44 . Recent evidence suggests that the former effect may be caused by inhibition of cytoplasmic ^PTP1B45 ; the cause of the latter is unclear. Currently, there are no tools to investigate the differential effects of spatially distinct PTP1B subpopulations on tumor-associated signaling events within the same cell. Photoswitchable variants of PTP1B represent such a tool.

Network biology. Signaling networks are typically represented as nodes (proteins) connected by lines (interactions) ⁴⁶ . Such diagrams capture the connectivity of biochemical relay systems but obscure spatial context—the ability of a single interaction to occur at multiple locations and potentially stimulate multiple signaling effects. The tools developed in this study will enable detailed studies of the role of spatial context in directing the propagation of signals through signaling biochemical networks; for example, understanding the role of PTP1B in cellular signaling (and processes relevant to tumorigenesis), and generally relevant to the study of any enzyme present in spatially distinct subpopulations within cells.

Generalization of the method to protein tyrosine phosphatases and kinases.

Two observations suggest that our framework for light control (i.e., attachment of the N-terminus of LOV2 to the C-terminal α-helix of the enzyme) is broadly applicable to PTPs and PTKs. (i) Structural alignments show that all PTPs possess, or with some mutations, the same allosteric communication network as PTP1B (Fig. 8A) ²³ . (ii) PTKs contain a C-terminal α-helix that is distal to their active site but is also able to modulate their catalytic activity (Fig. 8B) ⁶¹ .

We will evaluate the generality of our approach by constructing photoswitchable variants of striatal-enriched protein tyrosine phosphatase (STEP) and protein tyrosine kinase 6 (PTK6; Figure 8A). STEP is a neuron-specific phosphatase that is overactive in several neurological disorders, primarily Alzheimer's disease, schizophrenia, and drug ^{addiction62,63} . PTK6, which may act orthogonally to PTP1B in several signaling pathways, is expressed in approximately 70% of triple-negative breast cancers and promotes ^{metastasis50,64} . Photoswitchable variants of STEP and PTK6, both of which are present in multiple spatially distinct subpopulations within ^cells50,62 , will enable detailed study of their intracellular signaling roles, which remain poorly characterized.

For STEP and PTK6, we will develop and measure the substrate specificity of photoswitch chimeras by using several kinetic assays. For STEP, we will use assays similar to those employed with PTP1B. For PTK6, we will use the ADP-Glo kit developed by Promega, Inc. ⁶⁵ . This assay, which is compatible with any peptide substrate, converts the ADP generated by PTK-catalyzed peptide phosphorylation into a luminescent signal. For both enzymes, we will collect crystal structures of the best chimeras.

Exemplary photoswitch construct sequences for expression in mammalian cells or within an operon in microbial cells. In some embodiments, sequences can be optimized for microbial expression.

PPTP1B-LOV2, version 7.1 (T406A): DNA sequence SEQ ID NO: XX:

Protein sequence: SEQ ID NO: XX:

PTP1B-LOV2, version 7.1 (S286A): DNA sequence: SEQ ID NO: XX:

Protein sequence: SEQ ID NO: XX:

TCPTP-LOV2, the best version:

DNA sequence: SEQ ID NO: XX:

Underlined letters indicate sequences from PTP1B. Protein sequence: SEQ ID NO: XX:

TCPTP-LOV2 V2: DNA sequence: SEQ ID NO: XX:

Protein sequence: SEQ ID NO: XX:

FRET sensor. Forster resonance energy transfer (FRET) was considered for monitoring the activity of PTP1B in living cells. When treated with Src kinase, the sensor exhibited a 20% decrease in FRET signal (Figure 21B). Previous imaging studies have indicated that a 20% change in FRET is sufficient to monitor intracellular kinase ^{activity54-56} . In order to improve the spatial resolution in imaging studies, we will attempt to further optimize our sensor (and use it to measure the activity of PTP1B in vitro).

Exemplary FRET sensor: underlined mClover3-SH2 -linker - bold substrate - underlined and bold mRuby3 .

mClover3-mRuby3: DNA sequence: SEQ ID NO: XX:

Protein sequence: SEQ ID NO: XX:

Exemplary mammalian expression vectors are used to express photoswitch constructs in mammalian cells.

For insertion into mammalian expression vectors, such as lentiviral vectors, pAcGFP1-C1 (Clontech); PTP1B-LOV2 (above); promoter, such as CMV: SEQ ID NO:XX: GCAAATGGGCGGTAGGCGTGTACGGTGGGAGGTCTATATAAGCAGAGCTGGTTTAGTGAACCGTCAGATC; RBS, such as Kozak consensus translation start site: GCCACCATG; intragenic spacer, such as P2A: DNA sequence: SEQ ID NO:XX: GGCAGCGGCGCCACCAACTTCTCCCTGCTGAAGCAGGCCGGCGACGTGGAGGAGAACCCCGGCCCC; protein sequence: GSGATNFSLLKQAGDVEENPGP, etc.

Exemplary FRET sensors include: Promoter: same as above; RBS: same as above, and the like.

An exemplary FRET sensor was considered to avoid overlap between the excitation/emission wavelengths of LOV2 (455/495, which we noted was only weakly ^{fluorescent70} ) and our FRET pair (505/515 for Clover and 560/605 for mRuby2). While we still expected to see some crosstalk during imaging, previous three-color imaging ^studies71 indicated that it would not interfere with our ability to perform the experiments described in this section.

Contemplated embodiments include, but are not limited to, invadopodia formation and EGFR modulation.

Photoswitchable variants of PTP1B were considered to determine whether cytosolic PTP1B released from the ER by proteolysis is specifically responsible for regulating invadopodia formation, or whether ER-bound PTP1B may play a similar role. The formation of invadopodia—actin-rich protrusions that enable matrix degradation—facilitates cancer cell invasion and ^metastasis45 .

Both PTP1B and PTK6 regulate epidermal growth factor receptor (EGFR), a regulator of cell proliferation and migration that exhibits aberrant activity in many cancers and inflammatory diseases ^{51, 76.} We will perform a combined analysis of the synergistic contribution of PTP1B and PTK6 to EGFR regulation in different regions of the cell using PTP1B variants stimulated by red light and PTK6 variants stimulated by blue light (or vice versa).

References for Part I, Part II, and Part V are listed below and are incorporated herein by reference:

1.Wray,J.,Kalkan,T.,Gomez-Lopez,S.,Eckardt,D.,Cook,A.,Kemler,R.&Smith,A.Inhibition of glycogen synthase kinase-3alleviates Tcf3repression of the pluripotency network and increases embryonic stem cell resistance to differentiation. Nat. Cell Biol. 13, 838-45 (2011).

2. Wu, Y. I., Frey, D., Lungu, O. I., Jaehrig, A., Schlichting, I., Kuhlman, B. & Hahn, K. M. A genetically encoded photoactivatable Rac controls the motility of living cells. Nature 461, 104-108 (2009).

3. Liu, H., Wu, Y., Zhu, S., Liang, W., Wang, Z., Wang, Y., Lv, T., Yao, Y., Yuan, D. & Song, Y. PTP1B promotes cell proliferation and metastasis through activatingsrc and ERK1/2 in non-small cell lung cancer. Cancer Lett.359,218-225(2015).

4. Danial, N. N. & Korsmeyer, S. J. Cell Death: Critical Control Points. Cell 116, 205-219 (2004).

5. Johnson, T. O., Ermolieff, J. & Jirousek, M. R. Protein tyrosine phosphatase1B inhibitors for diabetes. Nat. Rev. Drug Discov. 1, 696-709 (2002).

6. Koren, S. & Fantus, I.G. Inhibition of the protein tyrosine phosphatasePTP1B: potential therapy for obesity, insulin resistance and type-2 diabetesmellitus. Best Pract. Res. Clin. Endocrinol. Metab. 21, 621-640 (2007).

7.Pike, K.a, Hutchins, A.P., Vinette, V., Theberge, J.-F., Sabbagh, L., Tremblay, M.L. & Miranda-Saavedra, D. Protein tyrosine phosphatase 1B isaregulator of the interleukin-10-induced transcriptional program inmacrophages.Sci.Signal.7,ra43(2014).

8. Rhee, I. & Veillette, a. Protein tyrosine phosphatases in lymphocyte activation and autoimmunity. Nat. Immunol. 13, 439-447 (2012).

9. Zhu, S., Bjorge, J.D. & Fujita, D.J. PTP1B contributes to the oncogenicproperties of colon cancer cells through Src activation. Cancer Res. 67, 10129-10137 (2007).

10. Volinsky, N. & Kholodenko, B. N. Complexity of receptor tyrosine kinase signal processing. Cold Spring Harb. Perspect. Biol. 5, (2013).

11.Kennedy, M.B.Signal-Processing Machines at the PostsynapticDensity.Science.290,750-754(2000).

12.Lee, H.K., Takamiya, K., Han, J.S., Man, H., Kim, C.H., Rumbaugh, G., Yu, S., Ding, L., He, C, Petralia, R.S., Wenthold, R.J. , Gallagher, M. & Huganir, R.L. Phosphorylation of the AMPA receptor GluR1 subunit is required for synaptic plasticity and retention of spatial memory. Cell 112, 631-643 (2003).

13. Bence, K. K., Delibegovic, M., Xue, B., Gorgun, C. Z., Hotamisligil, G. S., Neel, B. G. & Kahn, B. B. Neuronal PTP1B regulates body weight, adiposity and leptinaction. Nat. Med. 12, 917-24 (2006) .

14. Wu, P., Nielsen, T. E. & Clausen, M. H. FDA-approved small-molecule kinase inhibitors. Trends Pharmacol. Sci. 36, 422-439 (2015).

15. Repina, N. A., Rosenbloom, A., Mukherjee, A., Schaffer, D. V. & Kane, R. S. At Light Speed: Advances in Optogenetic Systems for Regulating Cell Signaling and Behavior. Annu. Rev. Chem. Biomol. Eng. 8, 13-39 (2017).

16.Gautier,A.,Gauron,C,Volovitch,M.,Bensimon,D.,Jullien,L.&Vriz,S.Howto control proteins with light in living systems.Nat.Chem.Biol.10,533-41(2014) .

17. Krauss, U., Lee, J., Benkovic, S. J. & Jaeger, K. E. LOVely enzymes-Towards engineering light-controllable biocatalysts. Microb. Biotechnol. 3, 15-23 (2010).

18.Dagliyan,O.,Tarnawski,M.,Chu,P.-H.,Shirvanyants,D.,Schlichting,I.,Dokholyan,N.V.&Hahn,K.M.Engineering extrinsic disorder to control proteinactivity in living cells.Science.354, 1441-1444(2016).

19. Zhou,

20. Lukyanov, K.a, Chudakov, D.M., Lukyanov, S. & Verkhusha, V.V. Photoactivatable fluorescent proteins. Nat. Rev. Mol. Cell Biol. 6, 885-890 (2005).

21.Rodriguez,E.A.,Campbell,R.E.,Lin,J.Y.,Lin,M.Z.,Miyawaki,A.,Palmer,A.E.,Shu,X.,Zhang,J.&Tsien,R.Y.The Growing and Glowing Toolbox of Fluorescentand Photoactive Proteins.Trends Biochem.Sci.42,111-129(2017).

22. Lessard, L., Stuible, M. & Tremblay, M.L. The two faces of PTP1B incancer. Biochim. Biophys. Acta-Proteins Proteomics 1804, 613-619 (2010).

23. Barr, A.J., Ugochukwu, E., Lee, W.H., King, O.N.F., Filippakopoulos, P., Alfano, I., Savitsky, P., Burgess-Brown, N.A., Mtiller, S. & Knapp, S. Large- ScaleStructural Analysis of the Classical Human Protein Tyrosine Phosphatome. Cell136,352-363(2009).

24. Hubbard, S. R. & Till, J. H. Protein tyrosine kinase structure and function. Annu. Rev. Biochem. 69, 373-398 (2000).

25. Zayner, J.P., Antoniou, C. & Sosnick, T.R. The amino-terminal helixmodulates light-activated conformational changes in AsLOV2. J. Mol. Biol. 419, 61-74 (2012).

26. Peter, E., Dick, B. & Baeurle, S.A. Mechanism of signal transduction of the LOV2-Ja photosensor from Avena sativa. Nat. Commun. 1, 122 (2010).

27.Kaberniuk,A.A.,Shemetov,A.A.&Verkhusha,V.V.A bacterialphytochrome-based optogenetic system controllable with near-infraredlight.Nat.Methods 13,1-15(2016).

28. Auldridge, M.E. & Forest, K.T. Bacterial phytochromes: more than meets the light. Crit. Rev. Biochem. Mol. Biol. 46, 67-88 (2011).

29. Anderie, I., Schulz, I. & Schmid, A. Characterization of the C-terminalER membrane anchor of PTP1B. Exp. Cell Res. 313, 3189-3197 (2007).

30. Tonks, N.K. & Muthuswamy, S.K.A Brake Becomes an Accelerator: PTP1B-A New Therapeutic Target for Breast Cancer. Cancer Cell 11, 214-216 (2007).

31. Krishnan, N. & Tonks, N. K. Anxious moments for the protein tyrosinephosphatase PTP1B. Trends Neurosci. 38, 462-465 (2015).

32.T raves,P.G.,Pardo,V.,Pimentel-Santillana,M.,Gonzalez-Rodrfguez,A.,Mojena,M.,Rico,D.,Montenegro,Y.,Cales,C,Martfn-Sanz,P ., Valverde, a M. & Bosca, L. Pivotal role of protein tyrosine phosphatase 1B (PTP1B) in themacrophage response to proinflammatory and anti-inflammatory challenge. CellDeath Dis. 5, e1125 (2014).

33.Matulka,K.,Lin,H.H.,Hrfbkova,H.,Uwanogho,D.,Dvorak,P.&Sun,Y.M.PTP1B is an effector of activin signaling and regulates neuralspecification of embryonic stem cells.Cell Stem Cell 13,706-719( 2013).

34.Cortesio, C.L., Chan, K.T., Perrin, B.J., Burton, N.O., Zhang, S., Zhang, Z.Y. & Huttenlocher, A. Calpain 2 and PTP1B function in a novel pathway with Srcto regulate invadopodia dynamics and breast cancer cell invasion. J.CellBiol.180,957-971(2008).

35.Wiesmann,C,Barr,K.J.,Kung,J.,Zhu,J.,Erlanson,D.A.,Shen,W.,Fahr,B.J.,Zhong,M.,Taylor,L.,Randal,M.,McDowell, R. S. & Hansen, S. K. Allostericinhibition of protein tyrosine phosphatase 1B. Nat. Struct. Mol. Biol. 11, 730-737 (2004).

36.Alonso,A.,Sasin,J.,Bottini,N.,Friedberg,I.,Friedberg,I.,Osterman,A.,Godzik,A.,Hunter,T.,Dixon,J.&Mustelin,T. Protein tyrosine phosphatases in the human genome. Ce//117,699-711(2004).

37. Haj, F.G., Verveer, P.J., Squire, A., Neel, B.G. & Bastiaens, P.I.H. Imaging sites of receptor dephosphorylation by PTP1B on the surface of theendoplasmic reticulum. Science 295, 1708-1711 (2002).

38.Romsicki,Y.,Reece,M.,Gauthier,J.Y.,Asante-Appiah,E.&Kennedy,B.P. Chem.279,12868-12875(2004).

39. Haj, F.G., Sabet, O., Kinkhabwala, A., Wimmer-Kleikamp, S., Roukos, V., Han, H.M., Grabenbauer, M., Bierbaum, M., Antony, C, Neel, B.G. & Bastiaens ,P.I.Regulation of signaling at regions of cell-cell contact byendoplasmic reticulum-boundprotein-tyrosine phosphatase 1B.PLoS One 7,(2012).

40. Soysal, S., Obermann, E. C, Gao, F., Oertli, D., Gillanders, W. E., Viehl, C. T. & Muenst, S. PTP1B expression is an independent positive prognostic factor in human breast cancer. Breast Cancer Res. Treat. 137,637-644(2013).

41. Zhang, S. & Zhang, Z. Y. PTP1B as a drug target: recent developments in PTP1B inhibitor discovery. Drug Discov. Today 12, 373-381 (2007).

42.Wu,C,Zhang,L,Bourne,P.A.,Reeder,J.E.,Di Sant&apos;Agnese,P.A.,Yao,J.L,Na,Y.&Huang,J.Protein tyrosine phosphatase PTP1B is involved in neuroendocrine differentiation of prostate cancer.Prostate 66,1124-1135(2006).

43.Lessard,L,DP,L,Deblois,G.,Begin,L,Hardy,S.,Mes-Masson,A.,Saad,F.,Trotman,L,Giguere,V.&Tremblay,M.PTP1B is an androgen receptor-regulatedphosphatase that promotes the progression of prostate cancer. Cancer Res. 72, 1529-1537(2012).

44.Dube,N.,Bourdeau,A.,Heinonen,K.M.,Cheng,A.,Loy,A.L.&Tremblay,M.L.Genetic ablation of protein tyrosine phosphatase 1B accelerateslymphomagenesis of p53-null mice through the regulation of B-cell development.Cancer Res .65,10088-10095(2005).

45. Weaver, A.M. Invadopodia: Specialized cell structures for cancerinvasion. Clin. Exp. Metastasis 23, 97-105 (2006).

46.Cui,Q.,Ma,Y.,Jaramillo,M.,Bari,H.,Awan,A.,Yang,S.,Zhang,S.,Liu,L.,Lu,M.,O'Connor -McCourt, M., Purisima, E. O. & Wang, E. A map of human cancer signaling. Mol. Syst. Biol. 3, 152 (2007).

47. Repina, N.A., Rosenbloom, A., Mukherjee, A., Schaffer, D.V. & Kane, R.S. At Light Speed: Advances in Optogenetic Systems for Regulating Cell Signaling and Behavior. Annu. Rev. Chem. Biomol. Eng. 8, 13-39 (2017).

48.Lee,J.,Natarajan,M.,Nashine,V.C,Socolich,M.,Vo,T.,Russ,W.P.,Benkovic,S.J.&Ranganathan,R.Surface sites for engineering allosteric controlin proteins.Science 322,438-442( 2008).

49. Qin, Z., Zhou, X., Pandey, N. R., Vecchiarelli, H. A., Stewart, C. A., Zhang, H.H.Chronic Stress Induces Anxiety via an Amygdalar Intracellular Cascadethat Impairs Endocannabinoid Signaling. Neuron 85,1319-1331(2015).

50.Fan,G.,Lin,G.,Lucito,R.&Tonks,N.K.Protein-tyrosine phosphatase 1Bantagonized signaling by insulin-like growth factor-1 receptor and kinaseBRK/PTK6 in ovarian cancer cells.J.Biol.Chem.288 ,24923-34(2013).

51. Eden, E. R., White, I. J., Tsapara, A. & Futter, C. E. Membrane contacts between endosomes and ER provide sites for PTP1B-epidermal growth factor receptor interaction. Nat. Cell Biol. 12, 267-72 (2010).

52. Arregui, C.O., Gonzalez, A., Burdisso, J.E. & Gonzalez Wusener, A.E. Protein tyrosine phosphatase PTP1B in cell adhesion and migration. CellAdh.Migr.7, 418-423 (2013).

53.Badran,A.H.,Guzov,V.M.,Huai,Q.,Kemp,M.M.,Vishwanath,P.,Kain,W.,Nance,A.M.,Evdokimov,A.,Moshiri,F.,Turner,K.H.,Wang,P .,Malvar,T.&Liu,D.R.Continuous evolution of Bacillus thuringiensis toxins overcomes insectresistance.Nature 533,58-63(2016).

54. Sato, M. & Umezawa, Y. in Cell Biol. Four-Volume Set 2, 325-328 (2006).

55. Ouyang, M., Sun, J., Chien, S. & Wang, Y. Determination of hierarchicalrelationship of Src and Rac at subcellular locations with FRET biosensors. Proc Natl Acad Sci USA 105, 14353-14358 (2008).

56.Ting,a Y.,Kain,K.H.,Klemke,R.L.&Tsien,R.Y.Genetically encoded fluorescent reporters of protein tyrosine kinase activities in living cells.Proc.Natl.Acad.Sci.U.S.A.98,15003-15008(2001).

57.Wiesmann,C,Barr,K.J.,Kung,J.,Zhu,J.,Erlanson,D.A.,Shen,W.,Fahr,B.J.,Zhong,M.,Taylor,L,Randal,M.,McDowell,R.S. & Hansen, S.K. Allostericinhibition of protein tyrosine phosphatase 1B. Nat. Struct. Mol. Biol. 11, 730-737 (2004).

58. Halavaty, A. S. & Moffat, K. N- and C-terminal flanking regions modulate light-induced signal transduction in the LOV2 domain of the blue light sensor phototropin 1 from Avena sativa. Biochemistry 46, 14001-14009 (2007).

59. Yao,

60. Choy, M.S., Li, Y., Machado, L.E.S.F., Kunze, M.B.A., Connors, C.R., Wei, X., Lindorff-Larson, K., Page, R. & Peti, W. Conformational Rigidity and ProteinDynamics at Distinct Timescales Regulate PTP1B Activity andAllostery.Mol.Cell 65,644-658(2017).

61. Hubbard, S. R. & Till, J. H. Protein tyrosine kinase structure and function. Annu. Rev. Biochem. 69, 373-398 (2000).

62. Zhang, Y., Kurup, P., Xu, J., Carty, N., Fernandez, S. M., Nygaard, H. B., Pittenger, C, Greengard, P., Strittmatter, S. M., Nairn, A. C. & Lombroso, P. J. Genetic reduction of striatal-enriched tyrosine phosphatase(STEP)reverses cognitiveand cellular deficits in an Alzheimer's disease mousemodel.Proc.Natl.Acad.Sci.107,19014-19019(2010).

63.He, R., Yu, Z., Zhang, R. & Zhang, Z. Protein tyrosine phosphatases aspotential therapeutic targets. Acta Pharmacol. Sin. 35, 1227-1246 (2014).

64.Ito,K.,Park,S.H.,Nayak,A.,Byerly,J.H.&Irie,H.Y.PTK6inhibitionsuppresses metastases of triple-negative breast cancer via SNAIL-dependent E-cadherin regulation.Cancer Res.76,4406-4417(2016) .

65. Zegzouti, H., Zdanovskaia, M., Hsiao, K. & Goueli, S.A. ADP-Glo: ABioluminescent and Homogeneous ADP Monitoring Assay for Kinases. Assay DrugDev. Technol. 7, 560-572 (2009).

66. Strickland, D., Yao,

67.Cosentino,C,Alberio,L.,Gazzarrini,S.,Aquila,M.,Romano,E.,Cermenati,S.,Zuccolini,P.,Petersen,J.,Beltrame,M.,Van Etten,J.L. ,Christie,J.M.,Thiel,G.&Moroni,A.Engineering of a light-gated potassium channel.Science.348,707-710(2015).

68.Krishnan,N.,Koveal,D.,Miller,D.H.,Xue,B.,Akshinthala,S.D.,Kragelj,J.,Jensen,M.R.,Gauss,C.-M.,Page,R.,Blackledge,M ., Muthuswamy, S. K., Peti, W. & Tonks, N. K. Targeting the disordered C terminus of PTP1B with an allostericinhibitor. Nat. Chem. Biol. 10, 558-566 (2014).

69. Piserchio, A., Cowburn, D. & Ghose, R. Expression and purification of Src-family kinases for solution NMRstudies. Methods Mol. Biol. 831, 111-131 (2012).

70.Van Stokkum,I.H.M.,Gauden,M.,Crosson,S.,Van Grondelle,R.,Moffat,K.&Kennis,J.T.M.The primary photophysics of the Avena sativa phototropin1LOV2 domain observed with time-resolved emission spectroscopy.Photochem.Photobiol .87,534-541(2011).

71. Rowland, A. A., Chitwood, P. J., Phillips, M. J. & Voeltz, G. K. ER contacts define the position and timing of endosome fission. Ce//159, 1027-1041 (2014).

72. Fehr, M., Lalonde, S., Lager, I., Wolff, M. W. & Frommer, W. B. In vivo imaging of the dynamics of glucose uptake in the cytosol of COS-7cells by fluorescent nanosensors. J. Biol. Chem. 278, 19127 -19133(2003).

73.Feigner,P.L.,Gadek,T.R.,Holm,M.,Roman,R.,Chan,H.W.,Wenz,M.,Northrop,J.P.,Ringold,G.M.&Danielsen,M.Lipofection: a highly efficient,lipid-mediated DNA -transfection procedure.Proc.Natl.Acad.Sci.U.S.A.84,7413-7(1987).

74.Gil-Parrado,S.,Fernandez-Montalvan,A.,Assfalg-Machleidt,I.,Popp,O.,Bestvater,F.,Holloschi,A.,Knoch,T.A.,Auerswald,E.A.,Welsh,K. ,Reed,J.C,Fritz,H.,Fuentes-Prior,P.,Spiess,E.,Salvesen,G.S.&Machleidt,W.lonomycin-activated calpain triggers apoptosis.A probable role for Bcl-2 family members.J.Biol.Chem. 277,27217-27226(2002).

75. Faeder, J.R., Blinov, M.L. & Hlavacek, W.S. Rule-based modeling of biochemical systems with BioNetGen. Methods Mol. Biol. 500, 113-167 (2009).

76. Tiganis, T., Bennett, A.M., Ravichandran, K.S. & Tonks, N.K. Epidermalgrowth factor receptor and the adapter protein p52Shc are specific substrates of T-cell protein tyrosine phosphatase. Mol. Cell. Biol. 18, 1622-34 ( 1998).

III. Construction and detection of genetically encoded systems for bioactive agents: microbial inhibitor screening system.

As described herein, several types of operons have been developed, each operon being used for a specific purpose, including but not limited to testing small molecules for their ability to inhibit, activate or otherwise regulate selected PTPs and/or PTKs; operons for testing the inhibitory, activating or modulatory effects of small molecules provided within cells on selected PTPs and/or PTKs; and evolving one or more proteins or small molecules of interest. More particularly, it is contemplated that gene operons for insertion using transfection and breeding techniques well known in the art are provided to provide microbial cells in which the activity of an enzyme of interest (e.g., protein tyrosine phosphatase 1B, a drug target for treating diabetes, obesity and cancer) is associated with (i) cell luminescence, (ii) cell fluorescence or (iii) cell growth. In some embodiments, such operons are modified for detecting and/or evolving biologically active metabolites. When modified and/or induced to construct various metabolites, cells will be used to detect metabolites that inhibit/activate proteins of interest (e.g., PTP1B).

These operons allow microbial cells containing the operons to be used to perform the following tasks: Detection of bioactive molecules and non-natural bioactive metabolites. When grown in the presence of a bioactive molecule that is (i) cell permeable and (ii) a small molecule capable of inhibiting a protein of interest (e.g., PTP1B), the cell will be able to detect that molecule. That is, it will exhibit a concentration-dependent response in luminescence, fluorescence, or growth. Many non-natural bioactive metabolites have useful pharmaceutical properties. Examples include paclitaxel and artemisinin, plant-derived terpenoids used to treat cancer and malaria, respectively. When the metabolic pathways responsible for making such natural metabolites are installed into microbial cells that also contain our operons, those cells will be able to detect metabolite-based bioactivities of interest (e.g., the ability to inhibit PTP1B).

When installed into microbial cells, the gene operon links the activity of an enzyme of interest (e.g., protein tyrosine phosphatase 1B, a drug target for the treatment of diabetes, obesity, and cancer) to (i) cell luminescence, (ii) cell fluorescence, or (iii) cell growth.

Detection and/or evolution of biologically active metabolites. When modified and/or induced to build various metabolites, cells will be able to detect metabolites that inhibit/activate a protein of interest (eg, PTP1B).

These operons allow microbial cells containing the operons to be used to perform the following tasks: Detection of bioactive molecules and non-natural bioactive metabolites. When grown in the presence of a bioactive molecule that is (i) cell permeable and (ii) a small molecule capable of inhibiting a protein of interest (e.g., PTP1B), the cell will be able to detect that molecule. That is, it will exhibit a concentration-dependent response in luminescence, fluorescence, or growth. Many non-natural bioactive metabolites have useful pharmaceutical properties. Examples include paclitaxel and artemisinin, plant-derived terpenoids used to treat cancer and malaria, respectively. When the metabolic pathways responsible for making such natural metabolites are installed into microbial cells that also contain our operons, those cells will be able to detect metabolite-based bioactivity (e.g., the ability to inhibit PTP1B).

In some embodiments, methods for improving evolutionary molecules can be derived from Moses et al., "Bioengineering of plant (tri) terpenoids: from metabolic engineering of plants to synthetic biology in vivo and invitro." New Phytologist, Vol. 200, No. 1, wherein the reference describes the synthesis of artemisinic acid, a precursor to the antimalarial drug artemisinin as a diterpene expressed in E. coli. In addition, enzyme engineering or directed evolution of terpenoid biosynthetic enzymes, such as engineered enzymes, is described, along with a discussion of enhancing the production of terpenoids in E. coli to accept non-natural substrates and catalyze regio- and stereo-specific reactions with efficiencies comparable to those of natural enzymes. In some embodiments, the method of evolving molecules can be improved from Badran et al., "Continuous evolution of Bacillus thuringiensis toxins overcomes insect resistance." Nature, Vol. 533: 58, 2016, wherein the reference describes phage-assisted continuous evolutionary selection, which rapidly evolves high-affinity protein-protein interactions, and applies the system to evolve variants of the Bt toxin Cry1Ac that bind to the cadherin-like receptor from the insect pest Trichoplusia ni (TnCAD) that is not naturally bound by wild-type Cry1Ac.

A. Protein evolution systems and evolved bioactive metabolites.

In some embodiments, the evolutionary molecular approach can be used to construct drug leads that can be readily synthesized in microbial hosts. It addresses the long-standing challenge of low-cost drug development by narrowing the molecular search space for lead discovery using a set of sophisticated biophysical tools and analytical methods, and by explicitly considering the accessibility of biosynthesis of therapeutic molecules. The approach is different from contemporary efforts to use microbial systems to synthesize clinically approved drugs and their precursors, and is unique in that it focuses on using biology to systematically build new molecules. This will accelerate the speed and reduce costs of drug development.

The development of a drug requires optimization of many of its pharmacological properties—affinity, absorption, distribution, metabolism, excretion, toxicology, pharmacokinetics, and pharmacodynamics. ¹ The first of these properties—protein-ligand binding affinity—usually determines whether the other properties are worth measuring or enhancing and therefore represents the property of a drug lead. ² Despite advances in computational chemistry and structural biology, the rational design of ligands that bind tightly to proteins—ligands, hereafter collectively referred to as inhibitors—remains exceptionally difficult ^{. 3} As a result, drug development often begins with the screening of large molecular libraries. ⁴ Once an inhibitor is discovered, it must be synthesized in quantities sufficient for subsequent analysis, optimization, formulation, and clinical evaluation.

The difficulties associated with developing protein inhibitors are particularly problematic for natural products. These molecules, which account for more than 50% of clinically approved drugs, often possess favorable pharmacological properties (e.g., membrane permeability) ⁵ . Unfortunately, their low natural potency—which prevents the extraction of detectable amounts from natural sources—and their chemical complexity—which complicates chemical synthesis—make the preparation of sufficient quantities for post-screening analysis time-consuming and expensive ⁶ .

In some embodiments, enzymes are contemplated for use in the construction of terpenoid inhibitors that can be synthesized in E. coli; this approach exploits the chemical diversity (and often favorable pharmacological properties) of natural products without the limitations of natural scarcity. In some embodiments, detailed biophysical studies of the molecular-level origins and thermodynamic basis of affinity and activity in protein-terpenoid interactions are included for the rapid construction of high affinity inhibitors. In some embodiments, the development of selective inhibitors of protein tyrosine phosphatase 1B (PTP1B), a target for the treatment of diabetes, obesity, and cancer, is contemplated, in part for the use of enzymes to evolve drug leads that are amenable to synthesis.

Structurally varied terpenoids with different affinities for the allosteric binding pocket of protein tyrosine phosphatase 1B (PTP1B).

Hypothesis. The results indicate that abietic acid, a mono-carboxylated variant of abietadiene, is an allosteric inhibitor of PTP1B. Derivatives or structural analogs of abietadiene that differ in stereochemistry, shape, size, and/or chemical functionality (including the position of carboxylation) may have different affinities for the allosteric binding pocket of PTP1B.

In some embodiments, (i) mutants of abietadiene synthase, cytochrome P450, and halogenase are considered for making structural variants of abietadiene, (ii) GC/MS to identify those variants, (iii) preparative HPLC and flash chromatography to separate them, and (iii) isothermal titration calorimetry to determine their free energy, enthalpy, and entropy of binding. In some embodiments, a panel of structurally varied inhibitors are considered that have (i) affinities that differ by 100-fold and/or (ii) enthalpy and entropy of binding that suggest alternative binding geometries.

To examine the molecular basis and thermodynamic origins of affinity and activity in enzyme-terpenoid interactions.

Hypothesis. Enzymes that bind, functionalize and/or synthesize terpenoids possess large non-polar binding pockets. We hypothesize that both (i) the affinity of the enzyme for terpenoids and (ii) the activity of the enzyme on terpenoids are determined by the general shape and hydration structure of their binding pockets rather than by the location of specific protein-terpenoid contacts.

In some embodiments, a sophisticated set of biophysical tools (isothermal titration calorimetry, X-ray crystallography, molecular dynamics (MD) simulations, and NMR spectroscopy) are contemplated for (i) determining how protein-ligand contacts, water rearrangements, and conformational constraints contribute to differences in affinity between terpenoid inhibitors, and (ii) developing a set of empirical relationships that predict how mutations in terpene synthases and terpene-functionalizing enzymes affect general properties (e.g., shape) of their products.

To evolve high-affinity terpenoid inhibitors of PTP1B.

Hypothesis. Mutants from secondary metabolism (e.g., terpene synthases/cytochrome P450s and halogenases) are highly promiscuous; a single mutation in or near their active sites can drastically alter their product properties. Mutagenesis of a small number (i.e., 2-4) of such enzymes, selected for their ability to synthesize and/or functionalize diterpenes, would enable the development of PTP1B inhibitors with submicromolar affinity.

In some embodiments, high affinity inhibitors of PTP1B are contemplated by pairing (i) high throughput methods for detecting inhibitors with (ii) site-saturation and random mutagenesis. For (i), we will develop four alternative fluorescence or growth-coupled assays to screen libraries of mutant pathways (and their respective products). For (ii), we use biological structural analysis and sequence alignment to identify potential residues that will generate enzymes with favorable product properties.

To identify structure-activity relationships that enable the evolution of terpenoid inhibitors of arbitrary protein targets.

Hypothesis. Proteins that interact with similar classes of molecules (either by binding or catalysis) have structurally similar binding pockets. Methods to assess these structural similarities—and their effects on enzyme activity—may allow the identification of enzymes that are capable of synthesizing inhibitors of any given protein.

In some embodiments, a biophysical framework is contemplated for using the crystal structure of a protein as a starting point to identify enzymes capable of synthesizing inhibitors of that protein. We will examine (and formalize) the structural relationships between (i) the active sites of enzymes for synthesizing allosteric inhibitors of PTP1B and (ii) the allosteric binding pocket of PTP1B, and we will validate these relationships by using them to identify—and then test—new enzymes capable of synthesizing inhibitors of PTP1B and (alone) undecadienyl diphosphate synthase, a target for treating antibiotic-resistant bacterial infections.

Diabetes, obesity and cancer.

Protein tyrosine phosphatase 1B (PTP1B) contributes to insulin resistance in type 2 ^diabetes7 , ^leptin resistance in obesity8, and tumor growth in breast, colorectal, and lung ^cancers9,11 . To date, the development of selective tight-binding inhibitors of PTP1B (i.e., for the treatment of diabetes, obesity, and cancer) has been hampered by the structure of its active site, in which polar residues restrict tight binding to charged, membrane-impermeable molecules, and in which the structure, similar to that of the active sites of other protein tyrosine phosphatases (PTPs), leads to off-target ^{interactions12,14} . In this proposal, we will construct selective inhibitors of PTP1B that bind to its C-terminal allosteric site, a largely nonpolar region that is not conserved among phosphatases. Previous screening of large molecular libraries has identified several ligands that bind to this site, but no clinically approved drugs ^have yet been produced16,13. Identification of new molecular alternatives—the challenge addressed by this proposal—remains a goal of efforts to develop selective PTP1 B inhibitory therapeutics.

Drug Development. The development of enzyme inhibitors—or leads—represents an expensive part of drug development; lead identification and optimization takes an average of 3 years and $250M to complete for each successful drug (-20-30% of the total time and cost to bring a drug to market) ¹⁷ . The technology developed in this proposal could accelerate the pace and reduce the cost of drug development by narrowing the molecular search space in lead discovery, by enabling the rapid construction of structurally variant leads (often referred to as "backups") ¹⁸ , and by facilitating the scale-up of molecular synthesis.

Molecular recognition. Hydrophobic interactions—energetically favorable associations of nonpolar species in aqueous solutions—account for, on average, 75% of the free energy of protein-ligand association ¹⁹ . Unfortunately, hydrophobic interactions between ligands and proteins—which vary greatly in rigidity, morphology, chemical functionality, and hydration structure—remain difficult to predict ²⁰ . This study uses detailed biophysical analysis and explicit water calculations to examine the thermodynamic basis of hydrophobic interactions between terpenoids and protein binding pockets. It will develop model systems—and corresponding conceptual frameworks—for studying hydrophobic interactions in the context of structurally varying protein-ligand complexes, accounting for them in the design of biosynthetic pathways, and exploiting them in the construction of new drug leads.

Biosynthesis of new natural products.

Synthetic biology offers a promising avenue for the discovery and production of natural products. When the metabolic machinery of an organism is installed into a genetically controlled production host (e.g., Saccharomyces cerevisiae or Escherichia coli), it is able to synthesize complex compounds at high titers (relative to the natural host). This approach has enabled the efficient production of pharmaceutically relevant metabolites from unculturable or low-yielding organisms ^{21, 22} but, unfortunately, requires significant investments in time and resources for pathway discovery and optimization; as a result, its use is often limited to low-throughput characterization of newly discovered gene clusters or to the production of known pharmaceutically relevant molecules (e.g., paclitaxel, artemisinin, or opioids) ²² - ²⁴ .

In some embodiments, strategies for constructing new molecular functions using synthetic biology are contemplated. It starts with a pathologically relevant protein target and engineers pathway enzymes to produce molecules that selectively inhibit the target. This approach will produce molecules that can be produced in microbial hosts without the need for a large amount of pathway optimization (which relies on enzymes that can be expressed by default); therefore, it will expand the use of synthetic biology to the generation of leads and backups. It cannot replace conventional methods for synthesizing complex natural products, but rather is a complementary strategy for constructing new compounds that enhance the efficiency of developing pharmaceutical preparations.

In the presence of a mutated metabolic pathway (e.g., a plant-based version of the terpenoid production pathway in which the terpene synthase has been mutated), our operon will enable the screening of a large number of metabolites for their ability to inhibit our protein of interest. Such a platform could be used to evolve metabolites with specific biological activities.

Detection and/or evolution of highly selective molecules. We have developed our operon version idea to detect molecules that inhibit one protein relative to highly similar proteins. Currently, screening for molecular selectivity remains very difficult.

Relative to some other systems for detecting small molecule inhibitors, the advantages of the methods and systems described herein include but are not limited to molecules that can detect the catalytic activity of regulating or changing the enzyme. Moreover, some embodiments of the system described herein allow detection of test molecules that change the enzyme activity by binding to anywhere on its surface. As an example, the detection of inhibitors that inactivate PTP1B by binding to its C-terminal allosteric site is considered. This binding event-it distorts the catalytic movement of the WPD ring-does not necessarily prevent enzyme-substrate association. On the contrary, U.S. Patent No. US6428951, which is incorporated herein by reference in its entirety, can detect molecules that prevent enzyme-substrate binding by competing substrate binding sites (i.e., active sites). As another example, as an embodiment, it is considered to detect molecules that activate the enzyme of interest. On the contrary, by quoting the U.S. Patent No. US6428951, which is incorporated herein by reference in its entirety, there is a method for detecting only molecules that prevent enzymes from binding to their substrates or otherwise change the affinity of enzymes to their substrates. As another example, it is considered as an embodiment to detect molecules that do not require enzymes and substrates to interact with any specific affinity, orientation or half-life. In contrast, U.S. Pat. No. 6,428,951, which is incorporated herein by reference in its entirety, requires that the enzyme and substrate bind to each other with an affinity and orientation that enables assembly of a split reporter. As a result, the enzyme may need to be modified. Instead, the inventors used a "substrate trapping" mutant of PTP1B to improve its affinity for the substrate domain.

As another example, some embodiments are able to detect inhibitors of wild-type enzymes. In contrast, Patent No. US6428951 to Tu S., which is incorporated herein by reference in its entirety, requires the enzyme to be fused to one half of a split reporter molecule.

In addition, the following two publications are examples of methods for detecting molecules that only destroy enzymes and substrates. Among other things, this feature is similar to the U.S. Patent No. US6428951, "Proteinfragment complementation assays for the detection of biological or drug interactions," which is incorporated herein by reference in its entirety. Publication date: January 31, 2008, on the contrary, the patent describes a protein-fragment complementation assay (PCA) based on high-throughput bacteria, wherein when two protein fragments derived from the enzyme dihydrofolate reductase (DHFR) are co-expressed as fusion molecules in Escherichia coli, interactions in the absence of inhibitors are observed, followed by concentration-dependent colony growth. The reference points out that PCA can be applicable to detecting the interactions of protein micromolecules, and provides examples, including complementary fragment fusions and bait fusion fragments. In fact, protein tyrosine phosphatase PTP1B is provided for detecting examples of enzyme-substrate interactions and for detecting examples of survival assays of protein-substrate interactions using aminoglycoside kinases (AK), examples of antibiotic resistance markers for PCA based on dominant selection of Escherichia coli. In addition, PCA is described as being applied to the identification of small molecule inhibitors of enzymes; natural products or small molecules from a library of compounds of potential therapeutic value; survival assays that can be used for library screening; for detecting endogenous DHFR inhibitors, such as rapamycin; and for protein-drug interactions. Under the control of a separate inducible or constitutive promoter with an interceding regions sequence, for example, derived from the mel operon, the expression of the PCA complementary fragment and the fused cDNA library/target gene can be assembled into a single operon on a single plasmid, or with polycistronic expression. PCA can be applied to the detection of protein-small molecule interactions. In this concept, two proteins are fused to the PCA complementary fragment, but the two proteins do not interact with each other. The interaction must be triggered by a third entity, which can be any molecule that will bind to both proteins at the same time or induce the interaction of two proteins by causing conformational changes in one or both partners. Moreover, the PCA strategy is exemplarily applied to protein engineering/evolution in bacteria to generate peptides or proteins with novel binding properties, which may have therapeutic value using phage display technology. An example of evolution produces a novel zipper sequence. Other examples of evolution producing endogenous toxins are described.

WO2004048549, Dep-1 Receptor Protein Tyrosine Phosphatase Interacting Proteins And Related Methods, published on June 10, 2004, and incorporated herein by reference in its entirety, describes a screening assay for inhibitors that alter the interaction between a PTP and a tyrosine-phosphorylated protein that is a substrate of the PTP, such as dephosphorylation of phosphatase-1 (DEP-1) by density enhancement of a DEP-1 substrate. DEP-1 polypeptides can be expressed in bacterial cells (including E. coli) under the control of an appropriate promoter, such as the E. coli arabinose operon (P _BAD or P _ARA ). This reference is similarly limited to the focus of U.S. Pat. No. US6428951, incorporated herein by reference in its entirety; it is capable of detecting molecules that disrupt the binding of a substrate to an enzyme, rather than molecules that modulate (i.e., enhance or attenuate) the activity of an enzyme.

Relative to some other systems for detecting small molecule inhibitors, the advantages of the methods and systems described herein include, but are not limited to, enabling metabolites that change the catalytic activity of the enzyme to evolve. Badran et al., "Continuous evolution of Bacillus thuringiensis toxins overcomes insect resistance". Nature, Vol 533: 58, 2016, incorporated herein by reference in its entirety, described in the technology; and the platform of continuous evolution in general has been used to evolve proteins with different affinities for other protein/peptide substrates. However, it has not been used to evolve enzymes that produce small molecules (i.e., metabolites) that change enzyme activity or protein-protein interaction strength.

Another advantage of the methods and systems described herein, relative to some other systems for detecting small molecule inhibitors, includes the discovery of metabolites with targeted biological activity but unknown structure (e.g., the ability to inhibit protein tyrosine phosphatase 1B). There are many inventions related to the production of terpenoids in E. coli or S. cerevisiae (e.g., Moses et al., "Bioengineering of plant (tri) terpenoids: from metabolic engineering of plants to synthetic biology in vivo and in vitro." New Phytologist, Vol. 200, No. 1), wherein the reference describes the synthesis of artemisinic acid, a precursor to the antimalarial drug artemisinin as a diterpene expressed in E. coli. In addition, along with discussions of enhancing the production of terpenoids in E. coli, enzyme engineering or directed evolution of terpenoid biosynthetic enzymes is described, such as engineering enzymes to accept non-natural substrates and react with catalytic regions and stereospecificities with efficiencies comparable to those of natural enzymes; in many cases, the metabolic pathways responsible for making these terpenoids are mutated to improve production levels. However, the use of biosensors (i.e., constructs that report on the concentration of various metabolites) has focused on the detection of specific intermediates (e.g., farnesyl pyrophosphate, a precursor to terpenoids) rather than biosensors for combining (i) mutagenesis of metabolic pathways and (ii) specific biological activities (e.g., the ability to inhibit PTP1B) for the discovery of new bioactive molecules (which may have unknown structures).

High-throughput metabolic engineering.

Microbial pathways are most effectively optimized using high-throughput screening. Unfortunately, currently, such screens are rare, and those that are available rely on signals (e.g., absorbance or fluorescence, association with product-specific transcription factors, or growth-permitting essential metabolites) that are difficult to extend to broad classes of molecules (e.g., those without obvious optical or metabolic properties) ²⁷ . The presented work developed a high-throughput screen for terpenoids with a targeted activity—the ability to inhibit PTP1B—rather than a targeted structure. These activity-focused screens can be broadly used to construct (i.e., evolve) new bioactive small molecules.

Identification of novel inhibitors: a starting point. We recently discovered that abietic acid, the major component of resin acids, is an allosteric inhibitor of PTP1B (Fig. 22). We will use the immediate precursor of this molecule, abietadiene, as a starting point for inhibitor design. Abietadiene has several properties that make it particularly compatible with our approach: (i) it can be synthesized in E. coli at a titer (200 mg/L) that allows purification, NMR analysis, and calorimetric ^studies28 , (ii) mutants of its related terpene synthase, abietadiene synthase, produce a range of hydrocarbon scaffolds that differ in stereochemistry, shape, and size (and still allow for analysis and purification) ^29-31 , and (iii) it—and similar molecules—can be functionalized by cytochrome P450 and ^{halogenases32-34} .

Metabolic Engineering. We have engineered strains of E. coli to produce abietadiene in titers (>150 mg/L) sufficient to allow analytical methods (i.e., GC/MS, ITC, and NMR). Our biosynthetic pathway has two required operons: MBIS, which converts (RJ-mevalonate to farnesyl pyrophosphate (FPP), and TS, which converts farnesyl pyrophosphate to abietadiene. When mevalonate is not included in the vector, an optional operon, MevT, which converts acetyl-CoA to (RJ-mevalonate), is necessary. ³⁵ Plasmids pMevT and pMBIS were developed by the Keasling Laboratory. ³⁶ Plasmid pTS, which contains abietadiene synthase (ABS) from Abies sempervirens, was developed as in Morrone ²⁸ with genes for geranylgeranyl diphosphate synthase from Ajikumar. ³⁷

Improved Inhibitors. We evaluated the ability of minor structural perturbations of abietadiene derivatives to produce improved inhibitors by comparing four structurally related (and commercially available) molecules: abietic acid, neoabietic acid, levopimaric acid, and dihydroabietic acid ( FIG. 16 ). Surprisingly, dihydroabietic acid was ten times more potent than abietic acid (K, ˜25 uM vs. 250 uM). Our ability to find improved inhibitors in this small screen suggests that the types of structural changes explored in this study—and the size of the molecular library generated—are likely to yield improved inhibitors.

Functionalization of abietadiene. We evaluated the ability of cytochrome P450 b m3 mutants to functionalize abietadiene-like molecules by installing five readily available mutants (G3, KSA-4, 9-10A, 139-3, and _J , which were engineered for activity against amorphadiene ³⁸ and steroid ³⁹ ) into our heterologous pathway. Three mutants produced hydroxylated and/or carboxylated products, yielding up to 0.3 mg/L of abietic acid ( FIG. 17 ). The abietadiene-functionalization activity of mutants initially engineered for other targets suggests that we will be able to develop mutants of P450 _b m3 with greater activity against abietadiene-like molecules.

Biological structure analysis. We have crystallized PTP1B in our laboratory, collected X-ray diffraction data in collaboration with Peter Zwart at Lawrence Berkeley National Lab (LBNL), and solved its crystal structure (Figure 17A inset). We have also co-crystallized PTP1B with abietic acid. We will analyze these crystals in late July (first available beam time).

Recently, in collaboration with Haribabu Arthanari of Harvard Medical School, we expressed ^N15- labeled PTP1B and used it to collect two- ^{dimensional1H} - ^15N HSQC spectra (Figure 17A main figure). The spectra include PTP1B bound to abietic acid and known inhibitors (separately); we are currently processing the data. Preliminary results (X-ray and NMR) suggest that biostructural studies of PTP1B bound to different inhibitors will be straightforward.

High Throughput Screening. Upon binding to inhibitors (competitive and allosteric), PTP1B exhibits a conformational change that quenches its tryptophan fluorescence (the basis of one of our four high throughput screens). FIG. 17B indicates that this quenching can be used to distinguish between inhibitory extracts (i.e., hexane overlays) from a strain of E. coli that produces abietadiene and non-inhibited extracts (i.e., those with catalytically inactive ABS) from a control strain. FIG. 17C indicates that this change can also be used to detect abietadiene at 50 uM (15 mg/L). Our ability to detect (i) abietadiene in culture extracts and (ii) low concentrations of abietadiene (i.e., ten times less potent than our abietadiene) suggests that we will be able to detect improved inhibitors of PTP1B, even if they are accompanied by a reduction in potency.

Structurally varied terpenoids with different affinities for the allosteric binding pocket are provided. This section describes the development of a panel of inhibitors with incremental differences in affinity resulting from systematic differences in structure. Goal (success metric): Inhibitors with a minimum of -15 structural variations have (i) 100-fold differences in affinity for PTP1B and/or (ii) enthalpy and entropy of binding that display alternative binding geometries.

Research plan. In the following sections, we use the enzyme to construct selective terpenoid inhibitors of PTP1B. This enzyme was the initial focus of our work because it is a therapeutic target for diabetes, obesity, and cancer, and can be easily expressed, crystallized, and ^assayed15 . Therefore, it serves as a pharmaceutically relevant model system with which to develop a general approach for enzymatic construction of drug leads.

Hypothesis of structural changes. In this section, we used promiscuous enzymes to construct terpenoids that differ in stereochemistry, shape, size, and chemical functionality. We believed that these modifications would affect the affinity of the ligands for PTP1B by altering: (i) their ability to participate in van der Waals interactions with nonpolar residues in the allosteric binding pocket (e.g., F280, L192, and F196), (ii) their ability to participate in direct or water-mediated hydrogen bonds with proximal polar residues (e.g., N193, E200, and E276), (iii) their ability to participate in halogen bonds with either set of residues, (iv) their effect on conformational constraints of the molecule; and (v) their ability to reorganize water during binding. This hypothesis, which is partially supported by Figure 16, drove the synthetic strategy described herein.

Stereochemistry, shape, and size. We will begin by using mutants of ABS to generate diterpenes that differ in stereochemistry and shape ( FIG. 18A ). ABS uses two active sites to sequentially catalyze type II (protonation-dependent) and type I (ionization-dependent) cyclizations of geranylgeranyl pyrophosphate (GGPP, C ₂ o) to abietadiene ²⁹ . Previous studies have indicated that amino acid substitutions in its active site can alter the stereochemistry or shape of its products ^{29 , 31} . We will use mutations (newly identified and previously identified) at positions that affect deprotonation, intramolecular protein transfer, or carbocation stability ( FIG. 8B ). After installing these mutants in E. coli , we will use GC/MS to search for new products (fragmentation tools such as MetFrag ⁴⁰ or ACD/MSFragmenter ⁴¹ will help identify new compounds).

We will generate terpenoids of varying sizes by using mutations that increase/decrease the volume of the ABS active site. Previous attempts to alter the substrate specificity of terpene ^{synthases42,43} have shown that such mutations can enhance activity towards farnesyl pyrophosphate (FPP, CI ₅ ) and farnesylgeranyl pyrophosphate (FGPP, C ₂ s). To synthesize FGPP, we will incorporate FGPP synthase previously expressed in E. ^coli44 .

We will isolate a subset of novel terpenoids with particularly high potency by using flash chromatography and HPLC (a task that has been established as feasible in several ^{studies28,31,45} ), and we will use ITC to measure the free energy (AG° _binding ), enthalpy (AH° _binding ), and entropy (-TAS° _binding ) of _binding to PTP1B. Differences in AG° _binding between ligands will reveal how structural changes affect binding strength. Differences in AH° and -TAS° _binding will reveal their effects on binding ^{geometry46,47} .

Hydroxylation and halogenation. For the three ligands selected in 6.1.2, we will use mutants of cytochrome P450 BM3 from Bacillus megaterium (P450 _b m3) and/or CYP720B4 from Sitka spruce (P450 ₇₂ o) to construct five variants with hydroxyl or carboxyl groups at different positions (Figures 19A and 19B). P450 _b m3 can hydroxylate a variety of different substrates, including terpenoids ^48. P450 ₇₂ o can carboxylate more than 20 diterpenes, including abietadiene ^49. Both enzymes can be expressed in E. coli ^{, 4A} .

We will work with several sets of mutations: for P450 _bm 3, we will use (i) three that allow stereoselective hydroxylation of sesquiterpenes and diterpenes (V78A, F87A, and A328L) ⁵⁰ , (ii) five that are able to hydroxylate alkaloids and steroids (L75A, M177A, L181A, and L437A) ⁵¹ , and (iii) two that allow carboxylation of heteroaromatic compounds (F87V and A82F) ( FIG18D ) ⁵² . For P450 ₇₂ o, we will examine −10 similar mutations that might alter the position of oxidation. We will again screen each mutant in E. coli, isolate the products of interest, and analyze them using ITC.

For each of the two high-affinity oxidized ligands, we will construct six variants with bromide or iodide at different positions ( FIG. 18C ). These two halogens can participate in halogen bonding with oxygen, nitrogen, or sulfur acceptors in proteins ⁵³ and can bind small nonpolar declivities on their surfaces ⁵⁴ . The energy contributions associated with both interactions tend to increase from Br to I ^54,55 and are therefore amenable to systematic analysis (i.e., physical organic methods). To generate halogenated ligands, we will use mutants of the tryptophan 6-halogenase (SttH) from Streptomyces toxigenin and the vanadium haloperoxidase (VHPO) from A _{CARYOCHLORISMARINA} . These enzymes can introduce halogens (chloride, bromide, or iodide) into sp ² -hybridized carbons of alkaloids or terpenoids (before or after cyclization) ^56,57 . For each enzyme, we will examine several mutations known to alter regioselectivity (e.g., L460F, P461E, and P452T for SttH ⁵⁶ ) and 5-10 mutations that may alter binding Figure 19 Examples: (Figure 19A) carboxylation, (B) hydroxylation, (Figure 19E) orientation of the ligand. We will again screen (Figure 19C) and halogenated diterpenes. (Figures 19D-E) residues of each mutant in E. coli and use ITC to target mutagenesis in (Figure 19D) P450 _bm 3 and (Figure 19E) SttH.

IV. Evolution of high-affinity terpenoid inhibitors of PTP1B.

This section develops four high-throughput screens for rapidly evaluating the potency of PTP1B inhibitors, and uses those methods along with site-saturation and random mutagenesis to evolve new inhibitors. Goal: A set of evolved inhibitors with exceptionally high affinity ( _KD ^1 uM) and/or unpredictable structures (i.e., structures inconsistent with rational design).

Bioselection. A selection method (ie, growth-coupled screening) in which the survival of E. coli is linked to inhibitor potency would enable rapid screening of very large libraries of molecules (10 ¹⁰ ) ⁶⁶ . In this section, we develop this approach.

PTP1B catalyzes the dephosphorylation—and inactivation—of several cell surface receptors. We will use the tyrosine-containing regions of these receptors to construct an operon that links inhibition of PTP1B to cell growth. The operon will require six components ( FIG. 21A ): (i) a substrate domain (the tyrosine-containing region of the receptor) tethered to a DNA binding protein, (ii) a substrate recognition domain (a protein that binds the tyrosine-containing region after it is phosphorylated) tethered to a co-subunit of RNA polymerase, (iii) a tyrosine kinase, (iv) PTP1B, (v) an antibiotic resistance gene, and (vi) an operator for that gene. Using this system, a PTP1B inhibitor will enable the substrate and substrate recognition domain to bind, recruit RNA polymerase to the DNA, and transcribe the gene for antibiotic resistance. Previous groups have used similar operons to evolve protein-protein binding partners. Here, we took the additional steps of (i) using protein-protein interactions mediated by enzymes (PTP1B and kinases) and (ii) screening this interaction in the presence of a potential inhibitor of one of those enzymes.

We will develop our operon starting with a luminescence-based system, and as a final step we will add antibiotic resistance genes. In our preliminary work using the system optimized by Liu et al. ⁶⁷ , we obtained a ten-fold difference in Lux-based luminescence between strains expressing two binding partners and those expressing one binding partner (Figure 21E; arabinose induces expression of the second partner). We now plan to introduce—and test—different substrate domains, recognition domains, and kinases (eGFR and Src).

FRET sensor for PTP1B activity. High-throughput screening of PTP1B inhibition coupled to cellular fluorescence would enable rapid screening via fluorescence-activated cell sorting (FACS). This technique tends to produce more false positives than selection and limits libraries to 10 ⁷ -10 ⁸ in size, but it requires fewer heterologous ^genes27,66 .

For this strategy, we will use FRET (Förster resonance energy transfer) sensors that are commonly used to monitor kinase and phosphatase activity in mammalian ^cells68,69 . These sensors consist of a kinase substrate domain, a short flexible linker, and a phosphorylation recognition domain—all sandwiched between two fluorescent proteins. Phosphorylation of the substrate domain leads to its binding to the recognition domain, thereby inducing FRET between the two fluorescent proteins. In a PTP1B-compatible sensor, a PTP1B inhibitor will increase FRET ( FIG21B ). We have begun to develop such a sensor by trying different combinations of substrate domain, recognition domain, and kinase. (Note: FACS is capable of FRET-based ^{screening70,71} ).

FRET sensor for changes in PTP1B conformation. FACS-based screening where changes in cellular fluorescence result from binding-induced conformational changes in PTP1B would be less general than strategies 2 and 3 (which can be used for any kinase or phosphatase), but would require only one heterologous gene.

For this strategy, we will use a FRET experiment performed by the Tonks ^Group13 . These researchers sought to show that binding of trodusquemine to PTP1B causes the protein to become more compact. To do this, they attached a member of a FRET pair to each end of PTP1B (Figure 21C). Upon protein-ligand association, an increase in the FRET signal indicates that their ends are close to each other. We hypothesize that this construct can be used as a sensor to identify other molecules that bind to the allosteric site of PTP1B. We will begin by testing it with various known inhibitors (a step not taken by the Tonks group).

Binding-induced changes in tryptophan fluorescence of PTP1B. Screening in which inhibition of PTP1B is correlated with changes in tryptophan fluorescence ( FIG. 21D ) will enable rapid screening of medium-sized libraries (10 ³ -10 ⁴ ) in microtiter plates ²⁷ . Our use of binding-induced changes in tryptophan fluorescence is described in 5.6 . In future work, we plan to extend this approach to other protein tyrosine phosphatases, many of which are allosteric and have many tryptophans (e.g., SHP-2, the target of Noonan syndrome ⁷² ).

Mutagenesis. To evolve PTP1B inhibitors using our high-throughput screen, we will construct a library of mutant terpenoid pathways by using: (i) site-saturation mutagenesis (SSM; we will target binary combinations of sites) and (ii) error-prone PCR (ep-PCR).

For SSM, we will identify "plastic" residues that are likely to accommodate useful mutations by developing a function similar to Equation 1. This function scores residues based on their ability to accommodate mutations that affect the volume and hydration structure of the active site; S is a measure of the propensity of a residue to allow mutations, cr ² is the volume difference of residues in similar positions in other enzyme active sites,

_s＝ 4 ₊ RTW _(EQ1}

A^ _w is the difference in hydrophilicity of those residues, and _Nv and _Nhw are normalization factors. In our preliminary analysis of ABS, we successfully used Equation 1 (and structural/sequence information from taxene, γ-humulene, 5-selenine, and epi-isozizaene synthases) to identify residues ³¹ known to have mutations that produced new products (e.g., H348 of ABS). We note that previous attempts to identify plastic residues have scanned every site in the vicinity of substrate binding ⁷³ ; our approach will be unique in that it includes biophysical considerations from: (i) our study of the properties of the optimal ligands (6.2.1) and (ii) our study of the types of mutations that brought them about (6.2.2).

For library construction, we will explore mutating our pathways (i) enzyme by enzyme (e.g., ABS, then P450 _b m3, and then VttH) or (ii) randomly. The second approach allows us to obtain structures that are difficult to find using conventional methods of lead design.

To identify structure-activity relationships that enable the evolution of terpenoid inhibitors of arbitrary protein targets. This section develops a biophysical framework for using the crystal structure of a protein to identify enzymes capable of making inhibitors of that protein. Objective: Use this framework to identify—and then test—enzymes capable of synthesizing new inhibitors of PTP1B and (alone) undecadienyl diphosphate synthase (UPPS), a target of antibiotic-resistant bacterial infections.

Relationships between binding pockets. We will begin by determining how similarities in specific properties of binding pockets (e.g., volume, polarity, and shape) allow enzymes to synthesize, functionalize, and/or bind similar molecules. This effort will involve comparing the allosteric binding pocket of PTP1B to binding pockets (i.e., active sites) of enzymes involved in inhibitor synthesis. For these comparisons, we will construct two matrices: matrix A, in which each element (ay) represents the similarity of a specific property between binding pockets i and j ((0<aij<1, where 1 is high similarity)), and matrix B, in which each element (by) describes the ability of binding pockets i and j to bind similar molecules (0<by<1, where 1 indicates identical binding specificity). The rank of the matrix formed by the product of these two matrices (AB) will show the number of independent variables (i.e., active site properties) necessary to determine the functional compatibility of enzymes in a metabolic pathway; the eigenvalues will show the relative importance of the properties studied (described by matrix A).

We will construct matrix A using PyMol- and MD-based protein crystal structure analysis. We will construct matrix B by examining the binding of functionalized terpenoids and their precursors to each of the enzymes involved in terpenoid synthesis. The binding affinity of some of these ligand/protein combinations will be measured using ITC; most will be evaluated using docking calculations (OEDocking ⁷⁸ ).

The result of this section will be an equation similar to Equation 2, where J is a measure of the ability of the active site to synthesize terpenoids that bind to a specific binding pocket

J = w _v V + w _p P + WiL + w _w W (Equation 2); V, P, L, and W represent specific properties of the active site (volume, polarity, longest diameter, and shortest diameter); w represents a weighting factor. The final number of variables - and their respective weights - will be determined by the above analysis. In parameterizing the equation, we plan to examine different measures of the properties of the binding pocket (e.g., shape) and explore/develop different substrate treatment methods.

Checksum extension.

Identification of promising active site motifs for inhibitor synthesis will require searching available protein structural data. We will perform this search using PROBIS (probis.nih.gov ⁷⁹ ), an alignment-based platform that uses a given binding site to find similar binding sites on other proteins in the Protein Data Bank. PROBIS can identify binding pockets of similar shape (i.e., it detects similar types of amino acids) even when the protein folds surrounding those pockets are different.

First, we will use PROBIS-based searches to identify enzymes with active sites that have a certain level of structural similarity to either (i) the allosteric binding site of PTP1B or (ii) the active site of an enzyme that can synthesize PTP1B inhibitors (we will explore different thresholds. Using Equation 2, we will select enzymes with the most favorable active sites and test them using our inhibitor discovery platform).

We will assess the generality of our approach by attempting to construct inhibitors of UPPS, a protein known to bind terpenoids and polycyclic ^molecules.80 The structure-based search will use two starting points: (i) UPPS and (ii) mutants of ABS, P450 _b m3 or similar enzymes that our biophysical analysis has shown to produce UPPS inhibitors. We will again select a subset of enzymes to test using our platform.

References used for Sections III-IV:

1. Koh, H.-L, Yau, W.-P., Ong, P.-S. & Hegde, A. Current trends in modern pharmaceutical analysis for drug discovery. Drug Discov. Today 8, 889-897 (2003).

2. Whitesides, G. M. & Krishnamurthy, V. M. Designing ligands to bind proteins. Q. Rev. Biophys. 38, 385-395 (2005).

3. Olsson, T.S.G., Williams, M.a., Pitt, W.R. & Ladbury, J.E. The Thermodynamics of Protein-Ligand Interaction and Solvation: Insights for Ligand Design. J. Mol. Biol. 384, 1002-1017 (2008).

4. Welsch, M.E., Snyder, S.A. & Stockwell, B.R. Privileged scaffolds for library design and drug discovery. Curr. Opin. Chem. Biol. 14, 347-361 (2010).

5. Gershenzon, J. & Dudareva, N. The function of terpene natural products in the natural world. Nat. Chem. Biol. 3, 408-414 (2007).

6. Chang, M.C.Y. & Keasling, J.D. Production of isoprenoid pharmaceuticals by engineered microbes. Nat. Chem. Biol. 2, 674-681 (2006).

7. Johnson, T. O., Ermolieff, J. & Jirousek, M. R. Protein tyrosine phosphatase1B inhibitors for diabetes. Nat. Rev. Drug Discov. 1, 696-709 (2002).

8. Koren, S. & Fantus, I.G. Inhibition of the protein tyrosine phosphatasePTP1B: potential therapy for obesity, insulin resistance and type-2diabetesmellitus. Best Pract. Res. Clin. Endocrinol. Metab. 21, 621-640 (2007).

9. Soysal, S., Obermann, E. C, Gao, F., Oertli, D., Gillanders, W. E., Viehl, C. T. & Muenst, S. PTP1B expression is an independent positive prognostic factor in human breast cancer. Breast Cancer Res. Treat. 137,637-644(2013).

10. Tonks, N.K. & Muthuswamy, S.K.A Brake Becomes an Accelerator: PTP1B-A New Therapeutic Target for Breast Cancer. Cancer Cell 11, 214-216 (2007).

11. Lessard, L, Stuible, M. & Tremblay, M.L. The two faces of PTP1B incancer. Biochim. Biophys. Acta-Proteins Proteomics 1804, 613-619 (2010).

12. Zhang, S. & Zhang, Z. Y. PTP1B as a drug target: recent developments in PTP1B inhibitor discovery. Drug Discov. Today 12, 373-381 (2007).

13.Krishnan,N.,Koveal,D.,Miller,D.H.,Xue,B.,Akshinthala,S.D.,Kragelj,J.,Jensen,M.R.,Gauss,C.-M.,Page,R.,Blackledge,M ., Muthuswamy, S. K., Peti, W. & Tonks, N. K. Targeting the disordered C terminus of PTP1B with an allostericinhibitor. Nat. Chem. Biol. 10, 558-566 (2014).

14.Sun,J.P.,Fedorov,A.A.,Lee,S.Y.,Guo,X.L,Shen,K.,Lawrence,D.S.,Almo,S.C.&Zhang,Z.Y.Crystal structure of PTP1B complexed with a potent and selective bidentate inhibitor.J.Biol.Chem .278,12406-12414(2003).

15.Wiesmann,C,Barr,K.J.,Kung,J.,Zhu,J.,Erlanson,D.A.,Shen,W.,Fahr,B.J.,Zhong,M.,Taylor,L.,Randal,M.,McDowell, R. S. & Hansen, S. K. Allostericinhibition of protein tyrosine phosphatase 1B. Nat. Struct. Mol. Biol. 11, 730-737 (2004).

16. Krishnan, N. & Tonks, N. K. Anxious moments for the protein tyrosinephosphatase PTP1B. Trends Neurosci. 38, 462-465 (2015).

17. Hughes, J.P., Rees, S.S., Kalindjian, S.B. & Philpott, K.L. Principles of early drug discovery. Br. J. Pharmacol. 162, 1239-1249 (2011).

18.Kennedy, T. Managing the drug discovery/development interface. Drug Discov. Today 2, 436-444 (1997).

19. Snyder, P.W., Lockett, M.R., Moustakas, D.T. & Whitesides, G.M. Is it the shape of the cavity, or the shape of the water in the cavity? Eur.Phys.J.Spec.Top.223,853-891(2014).

20. Klebe, G. Applying thermodynamic profiling in lead finding and optimization. Nat. Rev. Drug Discov. 14, 95-110 (2015).

21. Chang, M.C.Y. & Keasling, J.D. Production of isoprenoid pharmaceuticals by engineered microbes. Nat. Chem. Biol. 2, 674-681 (2006).

22. George, K. W., Alonso-Gutierrez, J., Keasling, J. D. & Lee, T. S. IsoprenoidDrugs, Biofuels, and Chemicals - Artemisinin, Farnesene, and Beyond. Adv Biochem Eng Biotechnol 148, 355-389 (2014).

23. Cragg, G. M. & Newman, D. J. Natural products: A continuing source of novel drug leads. Biochim. Biophys. Acta-Gen. Subj. 1830, 3670-3695 (2013).

24. Galanie, S., Thodey, K., Trenchard, I.J., Filsinger Interrante, M. & Smolke, C.D. Complete biosynthesis of opioids in yeast. Science (80-.). 349, 1095-1100 (2015).

25. Govindarajan, S., Recabarren, R. & Goldstein, R.a. Estimating the total number of protein folds. Proteins 35, 408-414 (1999).

26.Atanasov,A.G.,Waltenberger,B.,Pferschy-Wenzig,E.M.,Linder,T.,Wawrosch,C,Uhrin,P.,Temml,V.,Wang,L.,Schwaiger,S.,Heiss,E.H., Rollinger,J.M.,Schuster,D.,Breuss,J.M.,Bochkov,V.,Mihovilovic,M.D.,Kopp,B.,Bauer,R.,Dirsch,V.M.&Stuppner,H.Discovery and resupply of pharmacologically active plant-derived natural products :A review.Biotechnol.Adv.33,1582-1614(2015).

27.Lauchli,R.,Rabe,K.S.,Kalbarczyk,K.Z.,Tata,A.,Heel,T.,Kitto,R.Z.&Arnold,F.H.High-throughput screening for terpene-synthase-cyclizationactivity and directed evolution of a terpene synthase.Angew .Chemie-Int.Ed.52,5571-5574(2013).

28. Morrone, D., Lowry, L., Determan, M.K., Hershey, D.M., Xu, M. & Peters, R.J. Increasing diterpene yield with a modular metabolic engineering system in E. coli: Comparison of MEV and MEP isoprenoid precursor pathway engineering. Appl.Microbiol.Biotechnol.85,1893-1906(2010).

29. Peters, R.J. & Croteau, R.B. Abietadiene synthase catalysis: mutational analysis of a prenyl diphosphate ionization-initiated cyclization and rearrangement. Proc. Natl. Acad. Sci. U.S.A. 99, 580-584 (2002).

30.Wilderman,P.R.&Peters,R.J.A single residue switch convertsabietadiene synthase into a pimaradiene specific cyclase.J.Am.Chem.Soc.129,15736-15737(2007).

31.Criswell,J.,Potter,K.,Shephard,F.,Beale,M.H.&Peters,R.J.A singleresidue change leads to a hydroxylated product from the class II diterpenecyclization catalyzed by abietadiene synthase.Org.Lett.14,5828-5831( 2012).

32. Fasan, R. Tuning P450 enzymes as oxidation catalysts. ACS Catal. 2, 647-666 (2012).

33.Hamberger,B.B.,Ohnishi,T.,Hamberger,B.B.,Seguin,A.&Bohlmann,J.Evolution of diterpene metabolism: Sitka spruce CYP720B4catalyzes multipleoxidations in resin acid biosynthesis of conifer defense againstinsects.Plant Physiol.157,1677-95( 2011).

34.Fujimori,D.G.&Walsh,C.T.What's new in enzymatic halogenations.Curr.Opin.Chem.Biol.11,553-560(2007).

35. Martin, V.J.J., Pitera, D.J., Withers, S.T., Newman, J.D. & Keasling, J.D. Engineering a mevalonate pathway in Escherichia coli for production of terpenoids. Nat. Biotechnol. 21, 796-802 (2003).

36. Zhang, F. & Keasling, J. Biosensors and their applications in microbialmetabolic engineering. Trends Microbiol. 19, 323-329 (2011).

37.Ajikumar,P.K.,Xiao,W.-H.,Tyo,K.E.J.,Wang,Y.,Simeon,F.,Leonard,E.,Mucha,O.,Phon,T.H.,Pfeifer,B.&Stephanopoulos,G. Isoprenoid pathway optimization for Taxol precursor overproduction in Escherichia coli.Science330,70-74(2010).

38. Dietrich, J.A., Yoshikuni, Y., Fisher, K.J., Woolard, F.X., Ockey, D., McPhee, D.J., Renninger, N.S., Chang, M.C.Y., Baker, D. & Keasling, J.D. production using engineered substrate-promiscuous P450BM3. ACS Chem. Biol. 4, 261-267 (2009).

39. Zhang, K., El Damaty, S. & Fasan, R. P450 fingerprinting method for rapid discovery of terpene hydroxylating P450 catalysts with diversified regioselectivity. J. Am. Chem. Soc. 133, 3242-3245 (2011).

40. Ruttkies, C, Schymanski, E. L, Wolf, S., Hollender, J. & Neumann, S. Met Fragrelaunched: Incorporating strategies beyond in silicofragmentation. J. Cheminform. 8, (2016).

41.Pelander,A.,Tyrkko,E.&Ojanpera,I.In silico methods for predictingmetabolism and mass fragmentation applied to quetiapine in liquidchromatography/time-of-flight mass spectrometry urine drug screening.RapidCommun.Mass Spectrom.23,506-514 2009.

42. Kampranis, S.C, loannidis, D., Purvis, A., Mahrez, W., Ninga, E., Katerelos, N.A., Anssour, S., Dunwell, J.M., Degenhardt, J., Makris, A.M., Goodenough, P.W.&Johnson,C.B.Rational conversion of substrate and product specificity in aSalvia monoterpene synthase:structural insights into the evolution of terpenesynthase function.Plant Cell 19,1994-2005(2007).

43. Huang, Q., Williams, H.J., Roessner, C.A. & Scott, A.I. Sesquiterpenes produced by truncated taxadiene synthase. Tetrahedron Lett. 41, 9701-9704 (2000).

44. Tachibana, A., Yano, Y., Otani, S., Nomura, N., Sako, Y. & Taniguchi, M. Novelprenyltransferase gene encoding farnesylgeranyl diphosphate synthase from ahyperthermophilic archaeon, Aeropyrum pernix. Molecular evolution with alteration in product specificity. Eur.J.Biochem.267,321-328(2000).

45. Jia, M., Potter, K. C. & Peters, R. J. Extreme promiscuity of a bacterialand a plant diterpene synthase enables combinatorial biosynthesis. Metab. Eng. 37, 24-34 (2016).

46.Fox,J.M.,Kang,K.K.,Sastry,M.,Sherman,W.,Sankaran,B.,Zwart,P.H.&Whitesides,G.M.Water-Restructuring Mutations Can Reverse the ThermodynamicSignature of Ligand Binding to Human Carbonic Anhydrase.Angew.ChemieInt .Ed.56,3833-3837(2017).

47.Krimmer,S.G.,Betz,M.,Heine,A.&Klebe,G.Methyl,ethyl,propyl,butyl:Futile but not for water,as the correlation of structure and thermodynamicsignature shows in a congeneric series of thermolysin inhibitors.ChemMedChem9 ,833-846(2014).

48. Jung, S. T., Lauchli, R. & Arnold, F. H. Cytochrome P450: Taming a wild typeenzyme. Curr. Opin. Biotechnol. 22, 809-817 (2011).

49.Hamberger,B.B.,Ohnishi,T.,Hamberger,B.B.,Seguin,A.&Bohlmann,J.Evolution of diterpene metabolism: Sitka spruce CYP720B4catalyzes multiple oxidations in resin acid biosynthesis of conifer defense againstinsects.Plant Physiol.157,1677-95( 2011).

50. Seifert, A., Vomund, S., Grohmann, K., Kriening, S., Urlacher, V. B., Laschat, S. & Pleiss, J. Rational design of a minimal and highly enriched CYP102A1mutant library with improved regio-,stereo -and chemoselectivity.ChemBioChem10,853-861(2009).

51. Lewis, J.C, Mantovani, S.M., Fu, Y., Snow, C.D., Komor, R.S., Wong, C.H. & Arnold, F.H. Combinatorial alanine substitution enables rapid optimization of cytochrome P450BM3 for selective hydroxylation of large substrates. ChemBioChem 11, 2502 -2505

52.Butler,C.F.,Peet,C,Mason,A.E.,Voice,M.W.,Leys,D.&Munro,A.W.Keymutations alter the cytochrome P450 BM3 conformational landscape and removeinherent substrate bias.J.Biol.Chem.288,25387-25399( 2013).

53. Auffinger, P., Hays, F.a, Westhof, E. & Ho, P.S. Halogen bonds inbiological molecules. Proc. Natl. Acad. Sci. U.S.A. 101, 16789-16794 (2004).

54. Fox, J., Kang, K., Sherman, W., Heroux, A., Sastry, G., Baghbanzadeh, M., Lockett, M. & Whitesides, G. Interactions between Hofmeister anions and the binding pocket of a protein .J.Am.Chem.Soc.137,3859-3866(2015).

55. Carter, M., Voth, A.R., Scholfield, M.R., Rummel, B., Sowers, L.C. & Ho, P.S. Enthalpy-entropy compensation in biomolecular halogen bonds measured in DNA junctions. Biochemistry 52, 4891-4903 (2013).

56.Shepherd, S.A., Menon, B.R.K., Fisk, H., Struck, A.-W., Levey, C, Leyes, D. & Micklefield, J.A Structure-Guided Switch in the Regioselectivity of a Tryptophan Halogenase. ChemBioChem 17,821-824( 2016).

57.Carter-Franklin,J.N.,Parrish,J.D.,Tschirret-Guth,R.A.,Little,R.D.&Butler,A.Vanadium haloperoxidase-catalyzed bromination and cyclization ofterpenes.J.Am.Chem.Soc.125,3688-3689(2003) .

58. Li, R., Chou, W. K. W., Himmelberger, J. A., Litwin, K. M., Harris, G. G., Cane, D. E. & Christianson, D. W. Reprogramming the chemodiversity of terpenoidcyclization by remolding the active site contour of epi-isozizaenesynthase. Biochemistry 53, 1155— 1168(2014).

59. Brown, S. & O'Connor, S.E. Halogenase Engineering for the Generation of New Natural Product Analogues. ChemBioChem 16, 2129-2135 (2015).

60. Steele, C.L., Crock, J., Bohlmann, J. & Croteau, R. Sesquiterpene Synthases from Grand Fir (Abies grandis). J. Biol. Chem. 273, 2078-2089 (1998).

61. Lu, Y. & Mei, L. Co-expression of P450 BM3 and glucose dehydrogenase by recombinant Escherichia coli and its application in an NADPH-dependent indigoproduction system. J. Ind. Microbiol. Biotechnol. 34, 247-253 (2007).

62. Tzeng, S.-R. & Kalodimos, C.G. Protein activity regulation by conformational entropy. Nature 488, 236-240 (2012).

63.Aramini,J.M.,Vorobiev,S.M.,Tuberty,L.M.,Janjua,H.,Campbell,E.T.,Seetharaman,J.,Su,M.,Huang,Y.J.,Acton,T.B.,Xiao,R.,Tong,L. &Montelione,G.T.TheRAS-Binding Domain of Human BRAF Protein Serine/Threonine Kinase ExhibitsAllosteric Conformational Changes upon Binding HRAS.Structure 23,1382-1393(2015).

64. Christianson, D.W. Structural biology and chemistry of the terpenoidcyclases. Chem. Rev. 106, 3412-3442 (2006).

65.O'Maille, P.E., Malone, A., Delias, N., Andes Hess, B., Smentek, L., Sheehan, I., Greenhagen, B.T., Chappell, J., Manning, G. & Noel, J.P. Quantitative exploration of the catalytic landscape separating divergent plants sequiterpene synthases. Nat. Chem. Biol. 4, 617-623 (2008).

66. Packer, M. S. & Liu, D. R. Methods for the directed evolution of proteins. Nat. Rev. Genet. 16, 379-394 (2015).

67.Badran,A.H.,Guzov,V.M.,Huai,Q.,Kemp,M.M.,Vishwanath,P.,Kain,W.,Nance,A.M.,Evdokimov,A.,Moshiri,F.,Turner,K.H.,Wang,P .,Malvar,T.&Liu,D.R.Continuous evolution of Bacillus thuringiensis toxins overcomes insectresistance.Nature 533,58-63(2016).

68.Ting,a Y.,Kain,K.H.,Klemke,R.L.&Tsien,R.Y.Genetically encodedfluorescent reporters of protein tyrosine kinase activities in living cells.Proc.Natl.Acad.Sci.U.S.A.98,15003-15008(2001).

69. Sato, M. & Umezawa, Y. in Cell Biol. Four-Volume Set 2, 325-328 (2006).

70.Dye, B.T. Flow cytometric analysis of CFP-YFP FRET as a marker for in vivo protein-protein interaction. Clin. Appl. Immunol. Rev. 5, 307-324 (2005).

71. Vereb, G., Nagy, P. & Szollosi, J. Flow cytometric FRET analysis of protein interaction. Methods Mol. Biol. 699, 371-92 (2011).

72.Chen,Y.-N.P.,LaMarche,M.J.,Chan,H.M.,Fekkes,P.,Garcia-Fortanet,J.,Acker,M.G.,Antonakos,B.,Chen,C.H.-T.,Chen,Z., Cooke,V.G.,Dobson,J.R.,Deng,Z.,Fei,F.,Firestone,B.,Fodor,M.,Fridrich,C,Gao,H.,Grunenfelder,D.,Hao,H.-X., Jacob,J.,Ho,S.,Hsiao,K.,Kang,Z.B.,Karki,R.,Kato,M.,Larrow,J.,La Bonte,L.R.,Lenoir,F.,Liu,G.,Liu,S.,Majumdar,D.,Meyer,M.J.,Palermo,M.,Perez,L.,Pu,M.,Price,E.,Quinn, C,Shakya,S.,Shultz,M.D.,Slisz,J.,Venkatesan,K.,Wang,P.,Warmuth,M.,Williams,S.,Yang,G.,Yuan,J.,Zhang,J. -H., Zhu, P., Ramsey, T., Keen, N.J., Sellers, W.R., Stams, T. & Fortin, P.D. Allosteric inhibition of SHP2phosphatase inhibits cancers driven by receptor tyrosine kinases. Nature 535,148-52(2016).

73. Yoshikuni, Y., Ferrin, T. E. & Keasling, J. D. Designed divergent evolution of enzyme function. Nature 440, 1078-1082 (2006).

technology.Microb.Cell Fact.11,14(2012).74.Gruet,A.,Longhi,S.&Bignon,C.One-step generation of error-prone PCRlibraries using

technology.Microb.Cell Fact.11,14(2012).

75. Dietrich, J. A., McKee, A. E. & Keasling, J. D. High-Throughput Metabolic Engineering: Advances in Small-Molecule Screening and Selection. Annu. Rev. Biochem. 79, 563-590 (2010).

76. Esvelt, K. M., Carlson, J. C. & Liu, D. R. A system for the continuous directed evolution of biomolecules. Nature 472, 499-503 (2011).

77. Feiler, C, Fisher, A.C, Boock, J.T., Marrichi, M.J., Wright, L., Schmidpeter, P.A.M., Blankenfeldt, W., Pavelka, M. & DeLisa, M.P. Directed Evolution of Mycobacterium tuberculosis (3-Lactamase Reveals Gatekeeper Residue ThatRegulates Antibiotic Resistance and Catalytic Efficiency. PLoS One 8, (2013).

78.Murphy,R.B.,Repasky,M.P.,Greenwood,J.R.,Tubert-Brohman,I.,Jerome,S.,Annabhimoju,R.,Boyles,N.A.,Schmitz,C.D.,Abel,R.,Farid,R.&Friesner, R.A.WScore:A flexible and accurate treatment of explicit water molecules inligand-receptor docking.J.Med.Chem.acs.jmedchem.6b00131(2016).doi:10.1021/acs.jmedchem.6b00131

79. Konc, J., Miller, B.T., Stular, T., Lesnik, S., Woodcock, H.L., Brooks, B.R. & Janezic, D. ProBiS-CHARMMing: Web Interface for Prediction and Optimization of Ligands in Protein Binding Sites.J. Chem.Inf.Model.55,2308-2314(2015).

80.Guo, R.-T.,Cao,R.,Liang,P.-H.,Ko,T.-P.,Chang,T.-H.,Hudock,M.P.,Jeng,W.-Y. ,Chen,C.K.-M.,Zhang,Y.,Song,Y.,Kuo,C.-J.,Yin,F.,Oldfield,E.&Wang,A.H.-J.Bisphosphonates target multiple sites in both cis-and trans-prenyltransferases.Proc.Natl.Acad.Sci.U.S.A.104,10022-10027(2007).

81.Teng,K.H.&Liang,P.H.Structures,mechanisms and inhibitors of undecaprenyl diphosphate synthase:A cis-prenyltransferase for bacterialpeptidoglycan biosynthesis.Bioorg.Chem.43,51-57(2012).

82.Zhu,W.,Zhang,Y.,Sinko,W.,Hensler,M.E.,Olson,J.,Molohon,K.J.,Lindert,S.,Cao,R.,Li,K.,Wang,K., Wang,Y.,Liu,Y.-L,Sankovsky,A.,de Oliveira,C.A.F.,Mitchell,D.a,Nizet,V.,McCammon,J.A.&Oldfield,E.Antibacterial drugleads targeting isoprenoid biosynthesis.Proc.Natl.Acad. Sci.U.S.A.110,123-8(2013).

83. Leonard, E., Ajikumar, P.K., Thayer, K., Xiao, W.-H., Mo, J.D., Tidor, B., Stephanopoulos, G. & Prather, K.L.J. Combining metabolic and protein engineering of a terpenoid biosynthetic pathway for overproduction and selectivity control.Proc.Natl.Acad.Sci.U.S.A.107,13654-13659(2010).

84.Ro,D.-K.,Paradise,E.M.,Ouellet,M.,Fisher,K.J.,Newman,K.L.,Ndungu,J.M.,Ho,K.A.,Eachus,R.A.,Ham,T.S.,Kirby,J.,Chang, M.C.Y., Withers, S.T., Shiba, Y., Sarpong, R. & Keasling, J.D. Production of the antimalarial drug precursor artemisinic acid in engineered yeast. Nature 440, 940-943 (2006).

V. Specific embodiments of bacterial systems for identifying small molecules that modulate enzyme activity.

As described herein, strains of E. coli were developed that include both: (i) a gene encoding system that links cell survival to regulated inhibition of a pathology-related enzyme (i.e., a drug target) from Homo sapiens (i.e., a "bacterial two-hybrid" or B2H system) and (ii) a pathway for metabolite biosynthesis. The gene encoding system described herein contains more genetic elements (e.g., it has more than one promoter) than traditionally constituted a single operon, but is sometimes referred to as an operon.

More specifically, as described herein, a host organism, such as E. coli, is transformed with up to four plasmids, including: a first plasmid (plasmid 1), an expression plasmid, which includes a gene encoding system that links inhibition of a target enzyme to cell survival, wherein the target enzyme can be selected for the purpose of identifying molecules that inhibit a specific target enzyme; a second plasmid (plasmid 2), an expression plasmid, which includes an operon that expresses at least some of the genes necessary for the product of a synthetic metabolic pathway, such as a methylhydroxylase from terpenoid biosynthesis of Saccharomyces cerevisiae for providing a terpenoid product compound; valerate-dependent pathway; a third plasmid (plasmid 3), an expression plasmid comprising at least one additional gene not present in plasmid (plasmid 2), e.g., a terpene synthase, such as ADS, GHS, ABS or TXS, for providing a desired product, e.g., a terpenoid product, such that when the host bacteria express plasmids 1 and 2, the desired product is not produced until the host bacteria express plasmid 3 for completing the pathway to the desired compound; and a fourth plasmid (plasmid 4), comprising additional gene components specific to strains of E. coli, e.g., the F-plasmid of S1030 (Addgene 105063).

Examples of plasmid 1 embodiments are shown in Figures 33A, 33B, 33D, 33E, 34, 35, 40A, 40B, 40C, 40D, etc.

In some embodiments, the strain of E. coli used as a host for transformation has a ΔrpoZ mutation that enables the system encoded by Plasmid 1 to control the expression of genes for antibiotic resistance.

In some embodiments, plasmid 2 and/or 3 constitutes a pathway for terpenoid biosynthesis. In some embodiments, plasmid 2 and/or 3 constitutes a pathway for alkaloid biosynthesis. In some embodiments, plasmid 2 and/or 3 constitutes a pathway for polyketide biosynthesis.

In some embodiments, plasmid 3 further comprises a GGPPS gene in combination with ABS or TXS. Examples of GGPPS genes provide terpene synthase genes, i.e., substrates for ABS or TXS. In some embodiments, the terpene synthase gene is a wild-type gene. In some preferred embodiments, the terpene synthase gene comprises a mutation for producing a terpene product variant, as described and shown herein. In some embodiments, plasmid 3 further comprises a gene for a functional enzyme, such as a terpene of a cytochrome P450.

In some preferred embodiments, plasmid 1 is under the control of a constitutive promoter. Thus, in some preferred embodiments, at least some of the genes that are part of an operon in plasmid 1 are constitutively expressed. In some preferred embodiments, at least some of the genes that are part of an operon in plasmid 1 are expressed when contacted with an inducible compound, i.e., under the control of an inducible promoter, such as when contacted with X-gal, the lacZ promoter is turned on, and at least some of the genes that are part of an operon in plasmid 1 are expressed.

In some preferred embodiments, plasmids 2 and 3 are under the control of inducible promoters. Thus, in some preferred embodiments, at least some of the genes, and in some cases the entire set of genes, contained in the metabolic pathway operon in plasmid 2 are expressed upon contact with an inducible compound. In some preferred embodiments, some of the genes expressed in plasmid 3 are under inducible control.

In some preferred embodiments, plasmid 4 is under the control of a constitutive promoter. Thus, in some embodiments, at least one gene in plasmid 4 is under the control of a constitutive promoter. In some embodiments, at least one gene in plasmid 4 is under the control of an inducible promoter.

In some preferred embodiments, the host bacteria undergo at least 2 rounds of transformation, for example, first, plasmids 1 and 2 are simultaneously transformed into a strain already carrying plasmid 4 (e.g., an S1030 strain that already includes the accessory plasmid), followed by transformation with plasmid 3. In some preferred embodiments, the host bacteria undergo at least 3 rounds of transformation, for example, first, plasmid 1 is transfected, then plasmid 2 is transfected, and then plasmid 3 is transfected.

In some preferred embodiments, each plasmid has an antibiotic resistance gene (or other type of selectable gene) for identifying bacteria that have been successfully transformed with that plasmid, i.e., the antibiotic resistance gene may be different for each plasmid. Thus, when the antibiotic resistance gene is expressed, the bacteria are resistant and are therefore able to replicate at a normal or near normal rate, rather than bacteria that stop replicating normally when exposed to the antibiotic.

Thus, as described herein, a laboratory strain of E. coli is engineered to include up to three types of expression plasmids by first transfecting with plasmid 1, then selecting transformants (growing colonies) on/in an antibiotic-containing medium in which non-transformants do not grow, then transfecting the transformants with plasmid 2 and selecting for double transformants, e.g., an antibiotic-containing medium that allows growth of the double transformants, and then transfecting the double transformants with plasmid 3 and selecting for triple transformants, e.g., an antibiotic-containing medium that allows growth of the triple transformants. In one embodiment, the triple transformants are grown in medium containing inducers for the inducible plasmids (2 and 3) in combination with three antibiotics to produce a product having at least some inhibitory activity against the selected enzyme of plasmid 1, which is produced by the enzyme provided by the enzyme combination expressed by plasmids 2 and 3.

In addition, as described herein, a laboratory strain of E. coli is engineered to include up to four types of expression plasmids by first simultaneously transforming the host cell with plasmids 1 and 2 into a strain already carrying plasmid 4, then selecting triple transformants (growing colonies) on/in a medium containing antibiotics in which non-transformants do not grow, and then further transforming the successful triple transformants with plasmid 3 and selecting quadruple transformants, e.g., a medium containing antibiotics that allows the growth of quadruple transformants. In one embodiment, the quadruple transformants are grown in a medium containing (i) inducers of inducible plasmids (2 and 3), (ii) metabolic precursors for metabolite biosynthesis, e.g., mevalonate, and (iii) five antibiotics (i.e., one for each plasmid and one under the control of the gene encoding system in plasmid 1) for producing a product having at least some inhibitory activity against an enzyme selected by plasmid 1, which is produced by a combination of enzymes expressed by plasmids 2 and 3.

In some embodiments, the terpenoid operon pathway intended for insertion into plasmid 2 or already inserted into plasmid 2 can be altered by exchanging a different gene for a terpene synthase; (i.e., in each row of Figure 36, the metabolic pathway is different in the identity of the gene for the terpene synthase; when ADS or TXS is present, GGPPS is also present).

For example, in Figures 41A, 41B, 41C, and 41D, we mutated (rather than swapped) individual genes of a metabolic pathway: for example, at least one mutation was induced in a gene encoding amorphadiene synthase. By doing so, we show that metabolic pathways can be mutated to generate pathway libraries, and these pathways can be screened to identify pathways that produce more potent PTP1B inhibitors than the unmutated parental pathway.

In summary, we provide evidence that (i) the B2H system (detection operon) and (ii) metabolic pathways for terpenoid biosynthesis can be combined in a host organism to identify genes involved in small molecule production and to evolve genes associated with small molecule production that could be inhibitors, enabling microbial synthesis of PTP1B inhibitors.

In a preferred embodiment, the small molecule products are derived from a common metabolic pathway (the mevalonate-dependent pathway of terpenoid biosynthesis from Saccharomyces cerevisiae) and a host organism (Escherichia coli). The small molecule products produced as described herein are contemplated for use as treatments for type 2 diabetes, obesity, and breast cancer, among other diseases.

Without being bound by theory, when the gene encoding system for detecting the activity of a specific test enzyme is located in the host bacteria, the constitutive promoter expresses the A part of the detection system (e.g., the detection operon). As long as the phosphatase (or other test enzyme) expressed by the A part is active, the expressed kinase, such as Src kinase, attaches a phosphate (P) group to the expressed second fusion protein including a substrate recognition domain (S) attached to a protein capable of recruiting RNA polymerase to DNA (e.g., the RP _ω subunit of RNA polymerase), and the phosphatase removes the phosphate group, so that few molecules of the phosphorylated fusion protein 2 remain bound to the fusion protein 1, and therefore, there is almost no complex formation between the fusion proteins 2 and 1 to initiate transcription of the gene of interest (GOI).

Thus, transcription of part B is turned off and expression of the GOI is low, as observed, for example, when the GOI is a luminescent protein, as long as the placZ inducible promoter is not induced. In this embodiment of the operon, the placZ inducible promoter is induced so as to allow expression of the gene of interest in the absence of the inhibitor when the inhibitor molecule is not tested.

However, in the presence of a small molecule that inhibits the phosphatase, either produced endogenously from the metabolic pathway carried by plasmids 2 and 3, or added to the growth medium, then excess phosphorylated fusion protein 2 within the substrate binding region attaches to the substrate recognition domain of fusion protein 1, and then when both bind to the operator and RB binding sites, the GOI is then expressed, indicating the presence of a phosphatase inhibitor.

For practical purposes, it does not matter which fusion protein has the DNA binding protein and which has the protein capable of recruiting RNA polymerase to DNA, as long as the DNA binding protein forms part of one fusion protein and the protein that recruits RNA polymerase forms part of the other fusion protein. See, for example, FIG. 40 , FIG. 10 .

E. coli DH10B was used for molecular cloning and for preliminary analysis of terpenoid production; E. coli s1030 ¹ was used for luminescence studies and for experiments involving terpenoid-mediated selection (e.g., molecular evolution); and E. coli Bl21 was used for experiments involving heterologous expression and subsequent protein purification. However, it is not intended that these E. coli strains be host bacterial strains. Indeed, any bacterial strain that supports expression of operons, DNA sequences, and plasmids as described herein can be used as a host bacterial strain.

In a preferred embodiment, the small molecule products are derived from a common metabolic pathway (the mevalonate-dependent pathway of terpenoid biosynthesis from Saccharomyces cerevisiae) and a host organism (Escherichia coli). The small molecule products produced as described herein are contemplated for use as treatments for type 2 diabetes, obesity, and breast cancer, among other diseases.

A. Bacterial two-hybrid (B2H) system for identification of microbially synthesizable inhibitors of PTP1B (operon).

In one embodiment, the B2H system is applied to genes that evolve to enable the microbial synthesis of (i) molecules that inhibit PTP1B and (ii) molecules that can be identified (i.e., structurally characterized) by standard analytical methods. In short, the B2H system links the inactivation of PTP1B to the expression of genes for antibiotic resistance. Accordingly, when a strain of E. coli (or other host bacteria) carries both (i) the B2H system and (ii) a metabolic pathway for terpenoid biosynthesis, it will survive in the presence of antibiotics when it produces terpenoids that inhibit PTP1B.

The bacterial two-hybrid (B2H) system as described herein includes an embodiment of an operon as described herein. The data shown on the left side of the graph in Figure 33D (i.e., p130cas [also known as Kras] and MidT substrate) are the same as those shown in Figure 29A with additional details added in view of the development to provide the B2H system.

We propose to use directed evolution to evolve new inhibitors. That is, we will manually introduce mutations into specific genes (or genomes) within metabolic pathways to generate a metabolic pathway library that can be screened with the B2H system. Figure 41A describes a general method for introducing mutations; Example C provides a very specific way represented by Figure 41A. In order to screen our library, we transform it into cells containing B2H, and we culture them on plates containing various concentrations of spectinomycin. The colonies formed on the plates with high concentrations of spectinomycin contain pathways that can generate molecules that activate the B2H system (i.e., inhibit PTP1B). This pathway will not evolve naturally on its own. Therefore, we can remove it from the first host cell and transform it into another E. coli strain to make a high concentration of inhibitor.

Embodiments of the systems described herein enable rapid identification of drug leads that can be readily synthesized in microbial hosts. It allows for simultaneous solution of two problems encountered during drug development, which are typically examined separately: 1) identification of leads and 2) subsequent synthesis of leads identified in 1).

The system described herein has at least five purposes:

1. Genes that enable identification of proteins that generate inhibitors of drug targets. In short, when the pathway for terpenoid biosynthesis generates target-inhibiting molecules, cells survive high antibiotic concentrations. By swapping genes for terpenoid synthesis and/or functional enzymes, we can identify genes for enzymes that build such inhibitors.

2. Enables the construction of novel - and potentially non-natural - inhibitors. By mutating pathways for terpenoid biosynthesis, we can generate pathways that confer survival at high antibiotic concentrations. These pathways contain mutated (i.e., non-natural) genes and thus can generate inhibitor molecules not found in nature.

3. Ability to construct inhibitors that overcome drug resistance. In short, after constructing a strain that produces a target-inhibiting molecule, we can proceed with two steps: (i) We can mutate the drug target until it becomes resistant to the inhibitor. (ii) We can mutate the metabolic pathway until it generates an inhibitor of the mutated drug target. In this way, we can (i) predict drug-resistance mutations and (ii) address those mutations by generating new inhibitors that overcome them.

4. Inhibitors of protein tyrosine kinases can be constructed. Using a selection strategy similar to that described in 3.ii, we can mutate a metabolic pathway until it generates an inhibitor of Src kinase.

B. A genetically encoded system that links inhibition of protein tyrosine phosphatases to cell survival.

In a preferred embodiment, as described herein, a gene encoding system is developed and used to detect the presence of a selected enzyme, such as a small molecule inhibitor of the catalytic domain of a drug target enzyme, while allowing the survival of the host cell in the presence of a selective growth medium. In other words, when the gene encoding system is part of an expression plasmid in E. coli.

In one embodiment, an exemplary drug target enzyme is selected, such as a protein tyrosine phosphatase, protein tyrosine phosphatase 1B (PTP1B).

In one embodiment, the gene encoding system is part of an expression plasmid. In one embodiment, the sensing operon is operably linked to a constitutive promoter expressed in E. coli.

Figures 33A-E illustrate embodiments of a genetically encoded system for ligating enzyme activity to expression of a gene of interest (GOI). Error bars in Figures 33B-E represent standard deviation, n=3 biological replicates.

FIG33A illustrates an embodiment of a bacterial two-hybrid system for detecting phosphorylation-dependent protein-protein interactions. Components include (i) a substrate domain fused to the ω subunit of RNA polymerase (yellow), (ii) an SH2 domain fused to the 434 phage cI repressor (light blue), (iii) an operator of 434cI (dark green), (iv) a binding site for RNA polymerase (purple), (v) Src kinase, and (vi) PTP1B. Src-catalyzed phosphorylation of the substrate domain enables the substrate-SH2 interaction to activate transcription of the gene of interest (GOI, black). PTP1B-catalyzed dephosphorylation of the substrate domain prevents the interaction; inhibition of PTP1B re-enables it. FIG33B relates to an embodiment of the two-hybrid system from FIG33A that (i) lacks PTP1B and (ii) contains luxAB as GOI. We used inducible plasmids to increase expression of specific components; overexpression of Src enhanced luminescence. FIG. 33C relates to an embodiment of the two-hybrid system from FIG. 33A that (i) lacks both PTP1B and Src and (ii) includes a "superbinder" SH2 domain (SH2*, i.e., an SH2 domain with mutations that enhance its affinity for phosphopeptides), a variable substrate domain, and LuxAB as GOI. We used an inducible plasmid to increase the expression of Src; the luminescence increased most prominently for p130cas and MidT, indicating that Src acts on both substrate domains. FIG. 33D relates to an embodiment of the two-hybrid system from FIG. 33C with one of two substrates: p130cas or MidT. We used a second plasmid to overexpress (i) Src and PTP1B or (ii) Src and an inactive variant of PTP1B (C215S). The difference in luminescence between the systems containing PTP1B or PTP1B (C215S) was greatest for MidT, indicating that PTP1B acts on this substrate. Right: An optimized version of the two-hybrid system (bb030 as the RBS of PTP1B) is presented for reference. FIG. 33E shows the results of an exemplary growth-coupled assay using an optimized B2H comprising SH2*, midT substrate, an optimized promoter and ribosome binding site (bb034 of PTP1B) and SpecR as the GOI. The system is illustrated at the top of the figure. The exemplary growth results show that inactivation of PTP1B enables strains of E. coli to carry the system to survive high concentrations of spectinomycin (>250 μg/ml).

1. Sequential optimization of the two-hybrid system using LuxAB as the GOI.

Phase 1: We examined two different promoters of Src in a system lacking PTP1B. Phase 2: We examined two different ribosome binding sites (RBS) of Src in a system lacking PTP1B. Phase 3: We examined two different RBS of PTP1B in an intact system. Note: In Phases 1 and 2, the operon contained either the wild-type (WT) or non-phosphorylatable (mutant, Y/F) version of the substrate domain. In Phase 3, the operon contained either the wild-type (WT) or catalytically inactive (mutant, C215S) version of PTP1B. See, Figure 34.

FIG. 34 illustrates an exemplary experiment for optimizing the B2H system depicted in FIG. 33 .

2. Compare RB loci.

We grew strains of E. coli carrying bacterial two-hybrid versions of different RBSs (bb034 or bb030) containing PTP1B on various concentrations of spectinomycin (from left to right) and plated them on various concentrations of spectinomycin (from top to bottom). We used bb034 for one embodiment of the "optimized" two-hybrid system shown in Figure 33E. See, Figure 35.

FIG35 FIG3 illustrates an exemplary experiment for optimizing the B2H system depicted in FIG33 for use in a growth-coupled assay.

Rice, P., Longden, L. & Bleasby, A. EMBOSS: The European Molecular Biology Open Software Suite. Trends Genet. 16, 276–277 (2000).

C. Inhibition of terpenoid biosynthesis by PTP1B enables cell survival.

When pTS contained ADS or GHS, it did not contain GGPPS; when pTS contained ABS or TXS, it also contained GGPPS; ABS _D404A/D621A refers to a catalytically inactive variant of ABS; and B2H* contained PTP1B (C215S). ADS and, rarely, ABS were able to survive the presence of spectinomycin, indicating the ability of these to generate PTP1B inhibitors for terpene synthases.

Figure 36A-C Figure 4 | shows an illustration of an operon (Figure 36A) used to provide exemplary results during biosynthesis of terpenoids that inhibit PTP1B Figure 36B enabling cell survival Figure 36C.

36A-C FIG. 4 depicts exemplary metabolic pathways for the biosynthesis of terpenoids.

Figure 36A depicts plasmid loading pathways for terpenoid biosynthesis: (i) pMBIS, which carries the mevalonate-dependent isoprenoid pathway of Saccharomyces cerevisiae, converting mevalonate to isopentanyl pyrophosphate (IPP) and farnesyl pyrophosphate (FPP). (ii) pTS, which encodes a terpene synthase (TS) and, if necessary, a geranylgeranyl diphosphate synthase (GPPS), converting IPP and FPP to sesquiterpenes and/or diterpenes.

36B depicts exemplary terpene synthases: amorphadiene synthase (ADS) from Artemisia annua, gamma-humulene synthase (GHS) from Abies selengensis, abietadiene synthase (ABS) from Abies selengensis, and taxene synthase (TXS) from Taxus brevifolia.

Figure 36C shows the results of an exemplary growth-coupled assay of an E. coli strain containing both (i) an embodiment of an optimized bacterial two-hybrid (B2H) system (i.e., the B2H system from Figure 33E) and (ii) an embodiment of a pathway for terpenoid biosynthesis (i.e., the pathway from Figure 35A).

Briefly, we grew strains of E. coli carrying (i) the same pathway for the production of linear isoprenoid precursors and (ii) a different plasmid (pTS) encoding a terpene synthase. The pTS plasmid contained one of the following: (i) amorphadiene synthase (ADS) from Artemisia annua, (ii) gamma-humulene synthase (GHS) from Abies abies, (iii) abietadiene synthase (ABS) from Abies abies in operable combination with geranylgeranyl diphosphate synthase (GGPPS), (iv) taxene synthase (TXS) from Taxus brevifolia in operable combination with GGPPS, (v) an inactivated variant of ABS (i.e., ABS _xx , which corresponds to ABS _D404A/D621A ), or (vi) an L450Y mutant of GHS. After growth of these strains, we compared the ability of their products to inhibit PTP1B by performing the following steps: (i) we used a hexane overlay to extract the hydrophobic products (e.g., terpene-like products) from each culture medium, then we dried the products in a rotary evaporator, we dissolved the dried extracts in dimethyl sulfoxide (DMSO), and we measured the PTP1B-catalyzed hydrolysis of p-nitrophenyl phosphate (pNPP) in the presence and absence of DMSO containing the extracts. We note that the L450Y mutant of GHS was included in our analysis because the wild-type form of GHS does not allow B2H-mediated growth in the presence of antibiotics, but our preliminary data indicate that the L450Y mutant of GHS does not allow such growth. Accordingly, we hypothesized that the molecule produced by this mutant is a more potent inhibitor of PTP1B than the molecule produced by wild-type GHS. See, Figures 37A-C for demonstration of differential inhibition by structurally distinct terpenoids.

In examining Figures 37A-C, we observed a trend that extracts from strains containing terpene synthases that confer resistance to high concentrations of antibiotics (see Figure 36), with ADS and GHS _L450Y being more inhibitory than extracts from strains that do not confer resistance, such as TXS and ABS _xx . We note that strains containing ADS and GHS also comprise an optimized bacterial two-hybrid (B2H) system, but selection was not performed in the experiments for the product terpenoids of the experiments described by these figures.

Figures 37A-C provide exemplary analysis of the inhibitory effects of terpenoids produced by different strains of E. coli.

Figure 37A depicts the results of our analysis of the inhibitory effects of DMSO containing (i) no inhibitor and (ii) extracted compounds from the culture fermentation broth of strains containing ADS. Figure 37B depicts the results of our analysis of the inhibitory effects of DMSO containing (i) extracted compounds from the culture fermentation broth of strains containing GHS (gHUM) or (ii) extracted compounds from the culture fermentation broth of strains including the L450Y mutant of GHS. Figure 37C depicts the results of our analysis of the inhibitory effects of DMSO containing (i) no inhibitor, (ii) extracted compounds from the culture fermentation broth of strains containing ABS, (iii) extracted compounds from the culture fermentation broth of strains containing TXS, and (iv) extracted compounds from the culture fermentation broth of strains containing catalytically inactive variants of ABS.

Briefly, we grew strains of E. coli containing (i) an optimized bacterial two-hybrid system and (ii) a terpenoid pathway with a mutant γ-humulene synthase (GHS; 1 mutant/cell) on different concentrations of spectinomycin. Top: Product profile of strains with GHS mutants that confer survival at high antibiotic concentrations. See, Figure 38.

FIG. 38 shows an exemplary analysis of the product profile of GHS mutants capable of growing in the presence of spectinomycin.

In brief, we constructed a version of the bacterial two-hybrid system, which includes SH2*, midT substrate, optimized promoter and ribosome binding site, SpecR and alternative PTP: PTPN6 (e.g., SHP-1) and PTP1B ₄₀₅ (full-length version of PTP1B) catalytic domain. Note: These systems are identical to the B2H system depicted in Figure 33E, except that they only have one of the following PTP genes: PTP1B (as in Figure 33E), PTPN6 (different from Figure 33E) or full-length PTP1B. The inactivation of the catalytic domain of both PTPN6 and full-length PTP1B enables the strain of Escherichia coli carrying the corresponding operon to survive at high concentrations of spectinomycin (>400 μg/ml). In order to expand our operon to other PTPs, we plan to modify substrates, SH2 and/or kinase domains. See, Figure 39.

FIG. 39 Analysis of an exemplary B2H system linking inhibition of other PTPs to cell survival.

We also generated a version of the bacterial two-hybrid system that included SH2*, midT substrate, optimized promoter and ribosome binding site, SpecR and the catalytic domains of alternative PTPs: PTPN6 (e.g., SHP-1) and PTP1B ₄₀₅ (full-length version of PTP1B). Inactivation of the catalytic domains of PTPN6 and full-length PTP1B enabled strains of E. coli carrying the corresponding operons to survive high concentrations of spectinomycin (>400 μg/ml). To expand our operons to other PTPs, we plan to modify the substrate, SH2 and/or kinase domains.

Figures 40A-E depict exemplary embodiments of genetically encoded systems that link enzyme activity to expression of a gene of interest, and those embodiments are applied to (i) prediction of resistance mutations, (ii) construction of inhibitors to counteract resistance mutations, and (ii) evolution of kinase inhibitors.

FIG. 40A depicts an exemplary first step in examining potential resistance mutations. By evolving metabolic pathways to produce molecules that inhibit known drug targets (e.g., PTP1B); these molecules will allow expression of a gene of interest (GOI) that confers survival in the presence of selective pressure (e.g., the presence of spectinomycin, an antibiotic). FIG. 40B depicts an exemplary second step in examining potential resistance mutations. In a second strain of E. coli, we will replace the initial gene of interest with a second gene of interest (GOI2) that confers conditional toxicity (e.g., SacB, which converts sucrose to fructan, a toxic product); we will evolve the drug target to become resistant to the endogenous inhibitor while still retaining its activity. This mutant will prevent expression of the toxic gene. FIG. 40C depicts an exemplary third step in countering resistance mutations. In a third strain of E. coli, we will evolve metabolic pathways that produce molecules that inhibit the mutated drug target. In this way, we predict—and, through our second evolutionary pathway, address—mutations that may lead to resistance to terpenoid-based drugs. We note that Figures 40A-40C describe the use of our genetic encoding system to evolve inhibitors, but steps 2 and 3 can be used to predict mutations that allow resistance to endogenously supplied inhibitors, and then identify new endogenously supplied inhibitors that can counteract that resistance. Figure 40D depicts an exemplary genetic encoding system for detecting Src kinase inhibitors. In short, Src activity enables expression of a toxic gene (GOI2); then, inhibition of Src confers survival.

One embodiment of the construction of the B2H framework enables survival when PTP1B is active, i.e., when the activity of Src kinase is successfully abolished. In the absence of PTP1B, this construction can be used to evolve Src kinase inhibitors; such inhibitors can be used similarly for PTP1B by preventing phosphorylation of the substrate domain (as shown in Figure 40E). Src kinases are validated drug targets; tyrosine kinases are targets of more than 40 FDA-approved drugs.

Figure 40E illustrates one embodiment of the roof of principle for the B2H system described in Figure 40B. The system shown here includes two GOIs: SpecR and SacB. Expression of the GOIs confers survival in the presence of spectinomycin; expression of the GOIs results in toxicity in the presence of sucrose. The images depict the results of growth-coupled assays performed on strains of E. coli in the presence of various concentrations of sucrose. The strain carrying the active form of PTP1B (WT) grows better at high sucrose concentrations than the strain carrying the inactive form of PTP1B (C215S).

FIG. 41A depicts an exemplary strategy for the evolution of PTP1B inhibitors.

FIG41A depicts an exemplary structural analysis for identifying targets for mutagenesis in the active site of a terpene synthase. It shows the alignment of the class I active sites of ABS (grey, PDB entry 3s9v) and TXS (blue, PDB entry 3p5r) with the positions of the sites highlighted on ABS (red) for targeting site-saturation mutagenesis (SSM). Substrate analogs of TXS (yellow) appear for reference. FIG41B depicts an exemplary strategy for introducing diversity into a library of metabolic pathways: an iterative combination of SSM of key sites on a terpene synthase (as in a), error-prone PCR (ePCR) of the entire terpene synthase gene, SSM of sites on a terpene functionalizing enzyme (e.g., P450), and ePCR of the entire terpene functionalizing enzyme. FIG41C depicts an exemplary quantification of total terpenoids present in DMSO samples of extracts with various TS-containing strains. Briefly, we performed site-saturation mutagenesis at six sites on ADS (similar to the sites shown in Figure 41A); we plated the SSM library on agar plates containing different concentrations of spectinomycin; we picked colonies that grew on plates containing high concentration (800 μg/ml) spectinomycin and used each colony to inoculate a separate culture; we used a hexane overlay to extract the terpenoids secreted into the fermentation broth of each culture; we dried the hexane extracts in a rotary evaporator and resuspended the solids in DMSO; and we used GC-MS to quantify the total amount of terpenoids present in DMSO.

"ADS WT", "ADS F514E", "ADS F370L", "ADS G400A", "ADS G439A" and "ADS G400L" describe a mixture of molecules produced by a strain of E. coli carrying a mutant of amorphadiene synthase (ADS). The labels describe the mutants: "G439A" corresponds to a mutant of amorphadiene synthase in which glycine 439 has been mutated to alanine, etc. In future work, we plan to (i) purify different terpenoids from these mixtures, (ii) evaluate their inhibitory effects on PTP1B in vitro, (iii) analyze their inhibitory effects on other PTPs (especially TC-PTP and PTPN11) in vitro and (iv) analyze their effects on mammalian cells. See, Figure 41D.

Figure 41D depicts an exemplary analysis of the inhibitory effects of various extracts on PTP1B. Briefly, the figure shows the initial rate of PTP1B-catalyzed hydrolysis of p-nitrophenyl phosphate (pNPP) in the presence of the terpenoids quantified in Figure 41C. Two mutants of ADS (G439A and G400L) specifically generated potent PTP1B inhibitors.

FIG42 depicts an exemplary analysis of the link between B2H activation and cell survival. An exemplary strain of E. coli contains both (i) an optimized bacterial two-hybrid (B2H) system ( FIG33E ) and (ii) the terpenoid pathway depicted in FIG36A . Note: pTS includes GGPPS only when ABS or TXS are present; the "Y/F" operon corresponds to a B2H system in which the substrate domain cannot be phosphorylated. Survival at high concentrations of spectinomycin requires activation of the B2H system (i.e., phosphorylation of the substrate domain, a process promoted by inhibition of PTP1B).

FIG43 provides exemplary product profiles of strains of E. coli carrying various terpene synthases. For this figure, strains of E. coli carry (i) an optimized B2H system ( FIG33E ) and (ii) a terpenoid pathway ( FIG36A ). The pathways corresponding to each profile differ only in the composition of the pTS plasmids, which contain TXS (taxene synthase from Taxus brevifolia and geranylgeranyl diphosphate synthase from Taxus canadensis); GHS (γ-humulene synthase from Abies abies); ADS (amorphadiene synthase from Artemisia annua); ABS (abietenyl synthase from Abies abies and geranylgeranyl diphosphate synthase from Taxus canadensis); G400A (G400A mutant of amorphadiene synthase from Artemisia annua) and G439L (G439L mutant of amorphadiene synthase from Artemisia annua). Note that the two mutants of ADS produce products with characteristics different from those of the wild-type enzyme (ADS); our results indicate that the products generated by these two mutants are more inhibitory than those generated by the wild-type enzyme (Figure 41E).

D. Identification of sites for site-saturation mutagenesis (SSM).

The active sites of terpene synthases and cytochrome P450s contain clusters of amino acids that direct catalysis in two ways: (i) they control the conformational space available to the reacting substrates and (ii) they alter the organization of water around the substrates ^8–10 . We identified “plastic” residues that might modulate these properties in the class I active sites of terpene synthases by performing the following steps: (i) we aligned the crystal structure of ABS with that of TXS. (ii) we selected all residues within 8 angstroms of a substrate analogue of the class I active site of TXS (2-fluoro-geranylgeranyl diphosphate), and we identified a subset of sites that differed between ABS and TXS. (iii) we aligned the sequences of ABS,

是体积的变化、

是Hopp-Woods指数的变化以及n_v和n_HW是标准化因子(基于该研究中测量的最高变化)。(v)我们根据S排列每个位点并选择六个最高的评分位点。我们注意到：对于该分析，我们选择ABS和TXS，因为它们是结构上相似的具有晶体结构的酶(即，都具有α，β，和γ结构域)；我们选择GHS、DSS和EIS，因为它们已经显示，展示突变-响应产物特征。GHS, δ-selenine synthase (DSS) and epi-isozizaene synthase (EIS). (iv) We used Eq. S1 to score each site based on the variability of size and hydrophilicity of the five enzymes analyzed. In this equation,

is the change in volume,

is the change in the Hopp-Woods index and _nv and _nHW are normalization factors (based on the highest change measured in this study). (v) We ranked each site according to S and selected the six highest scoring sites. We note that: for this analysis, we selected ABS and TXS because they are structurally similar enzymes with crystal structures (i.e., both have α, β, and γ domains); we selected GHS, DSS, and EIS because they have been shown to exhibit mutation-response product characteristics.

Figures 44A-D provide exemplary structure- and sequence-based evidence supporting the extension of the B2H system to other protein tyrosine phosphatases (PTPs).

图44C证明了PTP1B和PTPN6的催化结构域的示例性序列比对(EMBOSS Needle¹)。图44D提供了PTP1B和TPPN6的催化结构域的示例性序列比较。序列共享34.1％序列同一性和53.5％序列相似性。概括来说，该图的结果指示我们的B2H系统可易于延伸至拥有催化结构域的PTP，该催化结构域(i)结构上类似于PTP1B的催化结构域(这里，我们定义结构相似性为这样的两个结构，当对齐时，具有≤

RMSD的RMSD，框架类似于通过PyMol的对齐功能使用的一个)，和/或(ii)序列类似于PTP1B的催化结构域(这里，我们定义序列相似性为≥34％序列同一性或≥53.5％序列相似性，如由EMBOSS Needle算法定义的)。为了鉴定能够调整P450_BM3的活性的“可塑”残基，我们进行类似于以上描述的方式：(i)我们使用突变体数据库¹¹(http://www.MuteinDB.org)来鉴定P450_BM3的功能变体中25个最常突变的位点。(ii)我们使用Eq.S1基于其不同突变体中大小和疏水性的变化性对每个位点评分。(iii)我们根据S排名每个位点和选择7个最高的评分位点。位点S1024基于S评分高，但是由于其位于P450还原酶结构域上被省略。Figure 44A provides an exemplary structural alignment of PTP1B and PTPN6, two PTPs that are compatible with the B2H system (see Figures 1e and 7 of Update A for evidence of compatibility). We used the alignment function of PyMol to align each structure of PTPN6 with the structure of the catalytic domain of PTP1B either (i) without ligand (3A5J) or (ii) with ligand (2F71). The alignment function performs sequence alignment followed by structural superposition, thus effectively aligning the catalytic domains of the two proteins. Figure 44B provides an exemplary structural comparison of PTP1B and PTPN6; the mean square root deviation (RMSD) of the aligned structures of PTP1B and PTPN6 ranges from 0.75 to 1.15.

Figure 44C demonstrates an exemplary sequence alignment of the catalytic domains of PTP1B and PTPN6 (EMBOSS Needle ¹ ). Figure 44D provides an exemplary sequence comparison of the catalytic domains of PTP1B and TPPN6. The sequences share 34.1% sequence identity and 53.5% sequence similarity. In summary, the results of this figure indicate that our B2H system can be easily extended to PTPs that possess a catalytic domain that is structurally similar to the catalytic domain of PTP1B (here, we define structural similarity as two structures that, when aligned, have ≤

RMSD of RMSD, framework similar to the one used by the alignment function of PyMol), and/or (ii) sequence similarity to the catalytic domain of PTP1B (here, we define sequence similarity as ≥34% sequence identity or ≥53.5% sequence similarity as defined by the EMBOSS Needle algorithm). To identify "plastic" residues that can adjust the activity of P450 _BM3 , we proceeded in a manner similar to that described above: (i) We used the Mutant ^Database11 (http://www.MuteinDB.org) to identify the 25 most frequently mutated sites in the functional variants of P450 _BM3 . (ii) We scored each site based on its variability in size and hydrophobicity among different mutants using Eq. S1. (iii) We ranked each site based on S and selected the 7 highest scoring sites. Site S1024 scored high based on S, but was omitted due to its location on the P450 reductase domain.

E. Exemplary purified products.

See Sections ^1–3 on Flash Chromatography and HPLC.

F. Exemplary test product concentration ranges.

We plan to incubate mammalian cells with 1-400 μM of the inhibitors; we will assess the biochemical impact of those inhibitors by using the assay described below.

G. Exemplary cell-based assays.

We will characterize the biological activity of newly developed inhibitors in at least two ways:

1. We will determine the effect of inhibitors on insulin receptor phosphorylation. Briefly, we will expose HepG2, Hela, Hek393t, MCF-7 and/or Cho-hIR cells to insulin shock in the presence and absence of inhibitors, and we will use Western blot and/or enzyme-linked immunosorbent assay (ELISA) to measure the effect of inhibitors on insulin receptor phosphorylation. In some embodiments, we can use cell-permeable PTP1B inhibitors to enhance insulin receptor phosphorylation.

2. We will examine the morphological and/or growth effects of inhibitors identified in the system described herein on cell models of HER2(+) and TN breast cancer.

Briefly, we will examine the relevance of inhibitors to HER2(+) breast cancer by evaluating their ability to inhibit the migration of HER2(+) BT474 and SKBR3 cells, but not HER2(-) MCF-7 and MDA-MB-231 cells. Next, we will examine the relevance of inhibitors to triple-negative breast cancer by performing viability and proliferation assays on a panel of TN cell lines (e.g., ATCC TCP-1002). All cell lines are available from ATCC (ATCC.org) and have been previously used to characterize potential therapies for HER2(+) and TN subtypes ^4,5 .

This is not meant to limit the pathway of terpenoid synthesis. Indeed, the alkaloid biosynthetic pathway is considered for identification,

An exemplary pathway for alkaloid biosynthesis consists of the following three modules (Nakagawa, A et al. A bacterial platform for fermentative production of plant alkaloids. Nat. Commun. (2011). doi: 10.1038/ncomms1327, incorporated herein by reference) (i) a first module capable of overexpressing our enzymes that overproduce L-tyrosine: TKT, PEPS, fbr-DAHPS and fbr-CM/PDH; (ii) a second module capable of expressing three enzymes necessary for the construction of dopamine and 3,4-DHPAA: TYR, DODC and MAO; and (iii) a third module capable of expressing four enzymes that construct (S) reticuline from 3,4-DHPAA and dopamine: NCS, 6OMT, CNMT and 4'OMT. The enzymes are as follows: TKT, transketolase (tktA, GenBank Accession No. X68025); PEPS, phosphoenolpyruvate (PEP) synthase (ppsA, GenBank Accession No. X59381); fbr-DAHPS, feedback-inhibition-resistant 3-deoxy-D-arabino-pimelate-7-phosphate synthase (aroGfbr, GenBank Accession No. J01591); fbr-CM/PDH, feedback-inhibition-resistant chorismate mutase/prephenate dehydrogenase (tyrAfbr, GenBank Accession No. M10431); TYR, tyrosinase from Streptomyces castaneoglobisporus (ScTYR containing tyrosinase and its adaptor protein, ORF378, G AY254101 and AY254102); DODC, DOPA decarboxylase from Pseudomonas putida (GenBank Accession No. AE015451); MAO, monoamine oxidase from Micrococcus luteus (GenBank Accession No. AB010716); NCS, norcoclaurine synthase from Coptis chinensis (GenBank Accession No. AB267399); 6OMT, norcoclaurine 6-O-methyltransferase from Coptis chinensis (GenBank Accession No. D29811); CNMT, coclaurine-N-methyltransferase from Coptis chinensis (GenBank Accession No. AB061863); 4′OMT, 3′-hydroxy-N-methylcoclaurine 4′-O-methyltransferase from Coptis chinensis (GenBank Accession No. D29812). We note that these three modules can be encoded by two plasmids.

References used in Section V are incorporated herein by reference in their entirety:

1. Jia, M., Potter, K. C. & Peters, R. J. Extreme promiscuity of a bacterialand a plant diterpene synthase enables combinatorial biosynthesis. Metab. Eng. 37, 24–34 (2016).

2.Criswell,J.,Potter,K.,Shephard,F.,Beale,M.H.&Peters,R.J.A singleresidue change leads to a hydroxylated product from the class II diterpenecyclization catalyzed by abietadiene synthase.Org.Lett.14,5828–5831( 2012).

3. Morrone, D. et al. Increasing diterpene yield with a modular metabolic engineering system in E. coli: Comparison of MEV and MEP isoprenoid precursor pathway engineering. Appl. Microbiol. Biotechnol. 85, 1893–1906 (2010).

4.Dagliyan,O.et al.Engineering extrinsic disorder to control proteinactivity in living cells.Science(80-.).354,1441–1444(2016).

5.Lehmann,B.D.et al.Identification of human triple-negative breastcancer subtypes and preclinical models for selection of targeted therapies.J.Clin.Invest.(2011).doi:10.1172/JCI45014

6. Dempke, W.C.M., Uciechowski, P., Fenchel, K. & Chevassut, T. Targeting SHP-1,2and SHIP Pathways: A novel strategy for cancer treatment? Oncology(Switzerland)(2018).doi:10.1159/000490106

7.Nakagawa,A.et al.A bacterial platform for fermentative production of plant alkaloids.Nat.Commun.(2011).doi:10.1038/ncomms1327

8. Christianson, D.W. Structural biology and chemistry of the terpenoidcyclases. Chem. Rev. 106, 3412–3442 (2006).

9. Fasan, R. Tuning P450 enzymes as oxidation catalysts. ACS Catalysis 2, 647–666 (2012).

10. Jung, S. T., Lauchli, R. & Arnold, F. H. Cytochrome P450: Taming a wild typeenzyme. Current Opinion in Biotechnology 22, 809–817 (2011).

11.Braun, A.et al.MuteinDB: The mutein database linking substrates, products and enzymatic reactions directly with genetic variants ofenzymes.Database(2012).doi:10.1093/database/bas028.

VI. Evolving optogenetic actuators: light-controlled switch constructs.

A. Optical control using red and infrared light.

Contemporary efforts to use light to control enzyme activity rely on at least two optogenetic actuators: LOV2, which has a terminal helix destabilized by blue light (-450 nm) ^2,18,48 , and Dronpa, which switches from a dimer to a monomer in response to green light (-500 nm) ¹⁹ . Unfortunately, blue and green light suffer from issues of phototoxicity, penetration depth, and spectral similarity that limit their use in signaling studies21. Therefore, in one embodiment, photoswitchable enzymes stimulated by red or infrared light are considered for development. These wavelengths have lower phototoxicity and greater penetration depth than blue and green ^{light20,21 ,} and would allow multicolor actuation with blue or green light.

B. Operons for evolving photoswitch constructs.

In one embodiment, an operon is contemplated that links the activity of PTP1B to cell growth. Briefly, the operon is based on the following control strategy (some additional details in FIG. 10 ): kinases stimulate the binding of two proteins, which in turn promotes transcription of essential genes; PTP1B inhibits the binding of the two proteins, thereby inhibiting transcription. The operon allows cells with a light-operated switch variant of PTP1B to grow faster in the presence of one light source than in the presence of another (e.g., 750 nm versus 650 nm). Differences in growth rates can identify functional chimeras. Initial experiments with an operon based on Lux luminescence (based on a system developed by Liu and colleagues ⁵³ ) showed a 20-fold difference in luminescence between strains expressing two model binding partners and strains expressing one ( FIG. 3E ). We will continue to develop the operon by adding protein-protein interactions regulated by PTPs and PTKs (see below).

This operon allows cells with a photoswitchable variant of PTP1B to grow faster in the presence of one light source than in the presence of another (e.g., 750 nm versus 650 nm). The difference in growth rate enables identification of functional chimeras. Initial experiments with a Lux-based luminescence operon (based on a system developed by Liu and colleagues ⁵³ ) showed a 20-fold difference in luminescence between strains expressing two model binding partners and those expressing one ( FIG3E ). We will continue to develop this operon by adding protein-protein interactions regulated by PTPs and PTKs.

FRET sensor. We will use Forster resonance energy transfer (FRET) to monitor the activity of PTP1B in living cells. Our preliminary sensor exhibited a 20% reduction in the FRET signal when treated with Src kinase (Figure 3F). Previous imaging studies indicate that a 20% change in FRET is sufficient to monitor intracellular kinase activity54 ^"56 . To enhance the spatial resolution in imaging studies, we will attempt to further optimize our sensor (and use it to measure the activity of PTP1B in vitro).

1. To evolve phosphatases and kinases regulated by red and infrared light.

This section uses directed evolution to construct enzymes that can be turned "on" and "off" with red and infrared light. We will know we are successful when we (i) construct a genetic operon that links the activity of PTP1B to antibiotic resistance, (ii) use this operon to construct PTP1 B-phytochrome chimeras that exhibit three-fold to ten-fold changes in activity in response to red and infrared light, and (iii) construct similar phytochrome chimeras for STEP and PTK6.

Hypothesis. Phytochrome proteins display global conformational changes when exposed to red and infrared light, ^27,28 but have so far eluded rational integration into photoswitchable enzymes. We hypothesized that genetic operons linking PTP or PTK activity to cell growth would be amenable to evolution of PTP- or PTK-phytochrome chimeras stimulated by red or infrared light.

Experimental Approach: We will construct an operon that links PTP1B inhibition to antibiotic resistance, and we will use the operon to evolve photoswitchable PTP1 B-phytochrome chimeras. This effort will involve (i) construction of a library of PTP1B-phytochrome chimeras that differ in linker composition and/or linker length, (ii) use of our operon to screen the library for functional mutants, (iii) kinetic and biostructural characterization of the most photoswitchable mutants, and (iv) extension of the approach to STEP and PTK6. This effort has two main goals: variants of PTP1B that are regulated by red and/or infrared light, and a general approach to extend optical control to new enzymes and different light wavelengths using directed evolution.

2. Development of a synthetic operon for the evolution of PTP1 B-phytochrome chimeras.

We will construct variants of PTP1B regulated by red and infrared light by attaching its C-terminal α-helix to the N-terminal α-helix of bacterial phytochrome protein 1 (BphP1) from Rhodopseudomonas palustris (Figure 9); this protein undergoes reversible conformational changes when exposed to 650nm and 750nm light. Phytochromes such as BphP1 are valuable for light control because they can actively switch between conformations (i.e., "on" and "off"). However, their structures are compatible with cage-based actuation (they do not undergo large-scale "unwinding"); therefore, they have been overlooked in previous efforts to develop light-operated switchable enzymes.

We will evolve a photoswitchable PTP1B-BphP1 chimera by using a genetic operon that links PTP1B activity to antibiotic resistance. This operon will consist of the following six components (Figure 10A-B): (i) a PTP1B substrate domain tethered to a DNA binding protein, (ii) a substrate recognition domain (i.e., substrate homology 2 domain or SH2) tethered to a subunit of RNA polymerase, (iii) a Src kinase (a kinase that can phosphorylate a variety of different substrates), (iv) PTP1B (or a potential photoswitchable variant of PTP1B), (v) a gene for antibiotic resistance, and (vi) an operator for that gene.

With this system, light-induced inactivation of PTP1B would enable transcription of genes for antibiotic resistance. Previous groups have used similar operons to evolve protein-protein binding partners (our system is based on an operon used by Liu et al. to evolve insecticidal proteins ⁵³ ); here, we take additional (new) steps: (i) using protein-protein interactions mediated by enzymes (phosphatases and kinases) and (ii) screening this interaction in the presence and absence of light.

We have begun developing our operon by using Lux-based luminescence as output. Preliminary results show that model protein-protein binding partners can cause a 20-fold change in luminescence (Figure 3E). We plan to swap out these binding partners with substrates and SH2 domains and test the new system with co-expressed PTP1B and Src kinase (which have some complementary activities and can be expressed in E. ^coli68,69 ).

The advantages of using an operator expressing a light-sensitive phosphatase include, but are not limited to, the ability to screen mutants of light-switchable enzymes in high throughput, and provide methods for screening enzyme libraries stimulated by them, see, Figure 5A for an example. In contrast, it has been shown that mutagenizing light-switchable enzymes can adjust (i.e., improve) their dynamic range (i.e., the ratio of dark-state activity to light-state activity); and some published studies, such as WO2011002977, Genetically Encoded Photomanipulation Of Protein And Peptide Activity, published on January 6, 2011, have proposed, but not demonstrated, that mutagenesis of protein photoswitches can enable light-switchable enzymes to be spectrally regulated. WO2011002977, provides a list of such sites that can be mutated to modify the flavin-binding pocket of LOV2 to accept flavins that absorb light at selectable wavelengths. However, their construct is described as the LOV2 domain of oat (Avena sativa) phototropin 1 (404-546), which includes a C-terminal helical extension Jα, in which Jα unwinds instead of the Aα helix described herein. Despite this, there are no available methods to perform high-throughput screening of mutants with modified binding pockets, and the invention described herein provides a platform for doing so. Furthermore, in contrast to WO2013016693, "Near-infrared light-activated proteins", published on January 31, 2013, the invention described herein provides such a platform for screening variants of potentially improved/modified light-operated switch proteins, such as plant phototropin 1 LOV2.

In addition, the method of screening libraries of enzymes is able to detect (i) molecules or (ii) photoswitchable domains that alter the activity of any enzyme, which can then modulate the affinity or protein-protein interaction-related results: protein tyrosine phosphatases (PTPs) and protein tyrosine kinases (PTKs) were confirmed. Moreover, proteases were considered to be added to the system as proteins.

C. directed evolution.

We will construct a library of PTP1B-BphP1 chimeras by pairing overlap extension PCR (oePCR) with error-prone PCR (epPCR). In particular, we will use oePCR to construct chimeras with varying linker lengths (here, we define the linker as a region of -20 residues that includes the C-terminal a-helix of PTP1B and the N-terminal a-helix of BphP1), and we will use epPCR to vary linker composition. Depending on the results of this initial library, we may extend the error-prone PCR into the BphP1 gene, but we will not mutate PTP1B beyond its C-terminal a-helix.

In the presence of small amounts of antibiotics (i.e., amounts that hinder the growth of E. coli), our gene operon will cause cells containing functional PTP1B-BphP1 chimeras to exhibit different growth rates under red light and infrared light. We will exploit these differences to identify cells carrying the photoswitch construct. Briefly, we will (i) generate two replicate plated colonies of cells, (ii) produce one under red light and one under infrared light ( FIG. 11A ), and (iii) select a subset of colonies (hot products) that show differential growth. We will further characterize our hot products by growing them in small-scale liquid media (e.g., 96-well plates with ˜1 ml/well; FIG. 11B ) under red and infrared light, and by sequencing the PTP1 B-BphP1 gene of the colonies that show the greatest difference in growth rate.

We will construct enzyme-phytochrome chimeras of STEP and PTK6 by pursuing two strategies: (i) we will replace PTP1B in our final PTP1B-BphP1 chimeras with either STEP or PTK6; this strategy will allow us to assess the modularity of our final designs, and (ii) we will use our operon-based approach to evolve functional STEP-BphP1 and PTK6-BphP1 chimeras; this strategy will allow us to assess the generalizability of our evolutionary approach.

The operators of the evolved STEP-BphP1 and PTK6-BphP1 chimeras will be very similar to the PTP IB-specific operators. For STEP, we will use STEP-specific substrates and SH2 domains (Src kinases with broad substrate specificity may have complementary activities on a subset of STEP substrates); for PTK6, we will use a recognition process that is inhibited by phosphorylation—not activated (here, we can use PTP1B _WT as the complementary enzyme).

D. Ways of expansion.

We will attempt to construct enzyme-phytochrome chimeras of STEP and PTK6 by pursuing two strategies: (i) we will replace PTP1B in our final PTP1B-BphP1 chimeras with STEP or PTK6; this strategy will allow us to assess the modularity of our final designs, and (ii) we will use our operon-based approach to evolve functional STEP-BphP1 and PTK6-BphP1 chimeras; this strategy will allow us to assess the generalizability of our evolutionary approach.

The operators of the evolved STEP-BphP1 and PTK6-BphP1 chimeras will be very similar to the PTP IB-specific operators. For STEP, we will use STEP-specific substrates and SH2 domains (Src kinases with broad substrate specificity may have complementary activities on a subset of STEP substrates); for PTK6, we will use a recognition process that is inhibited by phosphorylation—not activated (here, we can use PTP1BWT as a complementary enzyme).

F. Exemplary Considered Characterizations: Biophysical Characterization of Enzyme-Photochrome Chimeras.

We will examine the structural basis of light control in the most photoswitchable chimeras by using a subset of crystallographic and kinetic analyses. X-ray crystal structures will show how BphP1 affects the structures of PTP1B, STEP, and PTK6. Kinetic studies will show how BphP1 affects substrate specificity and binding affinity (or more specifically, Km, which is affected by binding affinity).

Table 1. Exemplary promoters

Table 2. Exemplary ribosome binding sites

Table 3. Exemplary protein sequences (including truncations)

Table 4. Exemplary terminators

Table 5. Exemplary DNA sequences (including truncations):

Abbreviations

PTP IB, protein tyrosine phosphatase IB; TC-PTP, T-cell protein tyrosine phosphatase; SHP2, protein tyrosine phosphatase non-receptor type 11; BBR, 3-(3,5-dibromo-4-hydroxy-benzoyl)-2-ethyl-benzofuran-6-sulfonic acid-(4-(thiazol-2-ylsulfamoyl)-phenyl)-amide; TCS401, 2-[(carboxycarbonyl)amino]-4,5,6,7-tetrahydro-thieno[2,3-c]pyridine-3-carboxylic acid hydrochloride; AA, abietic acid; SCA, statistical coupling analysis. _{PTP1B 1-435} , protein tyrosine phosphatase 1B (full length); SacB, levansucrase; GHS, γ-humulene synthase; ADS, amorphadiene synthase; ABS (or AgAs), abietadiene synthase; TXS, taxene synthase; PTPN5, protein tyrosine phosphatase non-receptor type 5; PTPN6, protein tyrosine phosphatase non-receptor type 6; PTPN11, protein tyrosine phosphatase non-receptor type 11; PTPN12, protein tyrosine phosphatase non-receptor type 12; PPTN22, protein tyrosine phosphatase non-receptor type 22; RpoZ, ω subunit of RNA polymerase; cI( or cI434), cI repressor protein from λ phage; Kras (or p130cas), p130cas phosphotyrosine substrate; MidT, phosphotyrosine substrate from hamster polyomavirus; EGFR substrate, phosphotyrosine substrate from epidermal growth factor receptor; Src, Src kinase; CDC37, Hsp90 co-chaperone Cdc37; MBP, maltose-binding protein; LuxAB, bacterial luciferase modules A and B; SpecR, spectinomycin resistance gene; GGPPS, geranylgeranyl diphosphate synthase; P450 (or P450 _BM3 ) cytochrome P450; LOV2, light-oxygen-voltage domain 2 from phototropin 1; BphP1, bacterial phytochrome; Galk, galactokinase.

Example

The following examples are provided to illustrate various embodiments of the present invention but should not be construed as limiting the scope of the present invention.

Statistical analysis of kinetic models. We evaluated the inhibition of four kinetic models as previously described (19). Briefly, we used the F test to compare the two-parameter hybrid model with several single-parameter models, and we used the Akaike information criterion (AIC or Ai) to compare the single-parameter models with each other. The hybrid model with p < 0.05 outperformed all single-parameter models, and the single-parameter model with Aj > 10 was inferior to the reference (i.e., "best fit") model.

Exemplary estimation of IC50. We estimated the half-maximal inhibitory concentration (IC50) of BBR by using a kinetic model to estimate the inhibitor concentration required to reduce the initial rate of PTP-catalyzed hydrolysis of 20 mM pNPP by 50%, and we used the MATLAB function "nlparci" to determine confidence intervals for those estimates (19).

All publications and patents mentioned in the above specification are incorporated herein by reference. Without departing from the scope and spirit of the present invention, various modifications and variations of the method and system of the present invention will be apparent to those skilled in the art. Although the present invention has been described in conjunction with specific preferred embodiments, it should be understood that the present invention as claimed should not be unduly limited to this specific embodiment. In fact, various modifications of the described modes of the present invention that are apparent to those skilled in the art for medicine, molecular biology, cell biology, genetics, statistics or related fields are intended to be within the scope of the appended claims.

Claims (15)

1. Methods for detecting operon systems using genetic codes, comprising, a. Provide, i. Inhibitor detection system comprising A first region of DNA and a second region of DNA, wherein: (A) The first region encoding of the DNA: (1) a first fusion protein under the control of a first promoter, wherein the first fusion protein comprises a Substrate Homology 2 domain linked to a DNA binding protein; (2) A second fusion protein under the control of said first promoter, wherein said second fusion protein comprises a protein kinase or protein phosphatase peptide substrate domain linked to a protein capable of recruiting RNA polymerase to DNA ; (3) Src protein kinase under the control of the second promoter; (4) a molecular chaperone under the control of said second promoter; (5) protein tyrosine phosphatase under the control of the second promoter, the protein tyrosine phosphatase comprising protein tyrosine phosphatase 1B (PTP1B), T-cell protein tyrosine phosphatase ( TC-PTP), protein tyrosine phosphatase non-receptor type 12 (PTPN12) or protein tyrosine phosphatase N6 (PTPN6); (B) the second region encoding of the DNA: (1) an operator for said DNA-binding protein; (2) a binding site for RNA polymerase, which serves as a third promoter; and (3) a reporter protein; and ii. an isoprenoid pathway DNA sequence encoding one or more components of the isoprenoid pathway under the control of a fourth promoter, iii. a synthetase DNA sequence encoding a synthetase under the control of the fifth promoter or said fourth promoter; and iv. various bacterial cells, and b. transfecting bacterial cells of said plurality of bacterial cells with said inhibitor detection system under conditions sufficient to express said reporter protein; c. transfecting said bacteria with said isoprenoid pathway DNA sequence under conditions sufficient to express said one or more components of said isoprenoid pathway; d. Transfecting the bacterium with the DNA sequence of the synthetase under conditions sufficient to express the synthetase; e. identifying said bacterial cells expressing said reporter protein among said plurality of bacterial cells, wherein expression of said reporter protein in said bacterial cells indicates that said bacterial cells comprise said protein synthesized by said synthetase Inhibitors of tyrosine phosphatases. 2. The method of claim 1, further comprising the step f. isolating an inhibitor molecule of said protein tyrosine phosphatase; and g. Treating mammalian cell culture with an inhibitor of said protein tyrosine phosphatase for reducing the activity of said protein tyrosine phosphatase. 3. The method of claim 2, wherein reducing the activity of the protein tyrosine phosphatase reduces growth of the mammalian cells of the mammalian cell culture. 4. The method of claim 1, wherein the protein tyrosine phosphatase is human PTP1B. 5. The method of claim 1, wherein the protein tyrosine phosphatase is wild-type. 6. The method of claim 1, wherein the protein tyrosine phosphatase has at least one mutation relative to the wild-type protein tyrosine phosphatase. 7. The method of claim 1, wherein the isoprenoid pathway DNA sequence comprises encoding mevalonate kinase, phosphomevalonate kinase, mevalonate pyrophosphate decarboxylase, isopentenyl The gene for pyrophosphate isomerase, farnesyl pyrophosphate synthase (ispA), or any combination thereof. 8. The method of claim 1, wherein the synthetase is a terpene synthase. 9. The method of claim 8, wherein the terpene synthase comprises amorphadiene synthase (ADS), gamma-humulene synthase (GHS), or a combination thereof. 10. The method according to claim 8, wherein said synthetase DNA sequence further encodes geranylgeranyl diphosphate synthase, and wherein said terpene synthase comprises abietene synthase, taxene synthase or a combination thereof. 11. The method of claim 8, wherein the terpene synthase is wild type. 12. The method of claim 1, wherein the inhibitor comprises a terpenoid or a structural variant of a terpenoid. 13. The method of claim 1, wherein expression of the reporter protein confers antibiotic resistance on the bacterial cell. 14. The method of claim 13, wherein the second region of the DNA encodes another reporter protein, wherein expression of the reporter protein and the other reporter protein confers the bacterium on two different Antibiotic resistance. 15. The method of claim 1 , further comprising providing said fourth promoter for said bacterial cells of said plurality of bacterial cells under conditions sufficient to contact said fourth promoter with an inducing molecule. Inducible molecules, wherein the fourth promoter is an inducible promoter.

Images